in vivo formation of hgse nanoparticles and hg
TRANSCRIPT
HAL Id: hal-03144114https://hal.archives-ouvertes.fr/hal-03144114
Submitted on 20 Mar 2021
HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.
In Vivo Formation of HgSe Nanoparticles andHg–Tetraselenolate Complex from Methylmercury in
Seabirds-Implications for the Hg–Se AntagonismAlain Manceau, Anne-Claire Gaillot, Pieter Glatzel, Yves Cherel, Paco
Bustamante
To cite this version:Alain Manceau, Anne-Claire Gaillot, Pieter Glatzel, Yves Cherel, Paco Bustamante. In Vivo For-mation of HgSe Nanoparticles and Hg–Tetraselenolate Complex from Methylmercury in Seabirds-Implications for the Hg–Se Antagonism. Environmental Science and Technology, American ChemicalSociety, 2021, 55 (3), pp.1515-1526. �10.1021/acs.est.0c06269�. �hal-03144114�
1
In vivo formation of HgSe nanoparticles and Hg-tetraselenolate complex from
methylmercury in seabird – Implications for the Hg-Se antagonism
Alain Manceau,*,† Anne-Claire Gaillot,‡ Pieter Glatzel,∥ Yves Cherel,§ Paco Bustamante⊥
†Univ. Grenoble Alpes, CNRS, ISTerre, 38000 Grenoble, France
‡Univ. Nantes, CNRS, Institut des Matériaux Jean Rouxel, IMN, 44000 Nantes, France
∥European Synchrotron Radiation Facility (ESRF), 71 Rue des Martyrs, 38000 Grenoble, France
§CEBC, CNRS, Univ. La Rochelle, 79360 Villiers-en-Bois, France.
⊥Univ. La Rochelle, CNRS, Littoral Environ. & Soc., LIENSs, 7266, La Rochelle, France
*Corresponding Author:
E-mail: [email protected]
Keywords: Mercury, bird, speciation, selenoprotein P, selenocysteine, XANES, EXAFS, STEM-
HAADF, STEM-EDX
ABSTRACT: In vivo and in vitro evidence for detoxification of methylmercury (MeHg) as insoluble
mercury selenide (HgSe) underlies the central paradigm that mercury exposure is not or little
hazardous when tissue Se is in molar excess (Se:Hg > 1). However, this hypothesis overlooks the
binding of Hg to selenoproteins, which lowers the amount of bioavailable Se that acts as a
detoxification reservoir for MeHg, thereby underestimating the toxicity of mercury. This question
was addressed by determining the chemical forms of Hg in various tissues of giant petrels
Macronectes spp. using a combination of high energy-resolution X-ray absorption near edge structure
(HR-XANES) and extended X-ray absorption fine structure (EXAFS) spectroscopy, and transmission
electron microscopy (HRTEM and STEM-HAADF) coupled to elemental mapping (EDX). Three
main Hg species were identified, a MeHg-cysteinate complex (MeHgCys), a four-coordinate
selenocysteinate complex (Hg(Sec)4), and a HgSe precipitate, together with a minor dicysteinate
complex Hg(Cys)2. The amount of HgSe decreases in the order liver > kidneys > brain = muscle, and
the amount of Hg(Sec)4 in the order muscle > kidneys > brain > liver. On the basis of biochemical
considerations and structural modeling, we hypothesize that Hg(Sec)4 is bound to the carboxy-
terminus domain of selenoprotein P (SelP) which contains 12 Sec residues. Structural flexibility
2
allows SelP to form multinuclear Hgx(Se,Sec)y complexes, which can be biomineralized to HgSe by
protein self-assembly. Because Hg(Sec)4 has a Se:Hg molar ratio of 4:1, this species depletes severely
the stock of bioavailable Se for selenoprotein synthesis and activity to one μg Se/g dry wet in the
muscle of several birds. This concentration is still relatively high because selenium is naturally
abundant in seawater, therefore it probably does not fall below the metabolic need for essential
selenium. However, this study shows that this may not be the case for terrestrial animals, and that
muscle may be the first tissue potentially injured by Hg toxicity.
INTRODUCTION
Being top predators in aquatic food webs, large seabirds are particularly exposed to methylmercury
(MeHg). The total mercury concentration commonly reaches several hundreds μg/g dry weight (dw)
in liver of old individuals, several tens μg/g dw in feathers, muscle, and kidneys, and several μg/g dw
in blood of piscivorous and scavenger seabirds from the Southern Ocean.1-8 Mercury concentrations
are positively correlated with selenium concentrations in the liver.9-14 Because mercury has a higher
affinity for selenium than for sulfur,15 the Hg-Se correlation is attributed chemically to the binding of
Hg(II) to Se(-II). The Hg-Se antagonism is well documented experimentally16-18 and in wildlife.19-21
Mercury selenide (HgSe) is the most common inorganic selenious form in biological tissues. HgSe
particles occur predominantly in the liver, and in lesser amounts in muscle, kidneys, brain, lung,
pancreas, and spleen of aquatic mammals (pinnipeds and cetaceans).22-26 Mercury sulfide (β-HgS), as
an admixture of primary HgSe, has been observed in the liver of the beluga whale Delphinapterus
leucas.27 Selenium and sulfur can also occur in solid solution Hg(Sx,Sey), as in the liver of the black-
footed albatross (Phoebastria nigripes).22 According to the best of the authors’ knowledge, no other
Hg form is known in seabirds.
The Hg-C bond of MeHg is cleaved readily by selenoamino acids under physiologically relevant
experimental conditions yielding HgSe as the end product.28 Hence, the biomineralization of
potentially inert and apparently nontoxic HgSe granules observed in wildlife is considered as the
main detoxification mechanism of MeHg. Consequently, the capacity for an organism to detoxify
MeHg depends on the Hg:Se molar ratio, which represents the fraction of Se bound to Hg.29-34
Methylmercury is more hazardous when the Hg level approaches or exceeds equimolar stoichiometry
with Se (i.e., Hg/Se ≥ 1),35, 36 to the point of inducing a conditioned deficiency of bioavailable Se for
selenoenzyme synthesis and activity.37, 38 Therefore, the difference of molar content between Se and
Hg ([Se]mol - [Hg]mol) is considered to reflect mercury sequestration and selenium depletion, and is
3
widely used to assess Hg exposure risk.12, 36, 39-41 Gajdosechova and coworkers26 cautioned that this
approach may be deceptive, however, because the cells may contain other selenious forms of Hg than
HgSe with a Hg:Se ratio different from one.
In addition to selenoamino acids, demethylation reactions can be mediated also by low molecular
weight seleniol molecules, such as selenoneine,42 and selenoproteins, such as selenoprotein P
(SelP).43 Selenoneine has been identified in fish,44, in cetacean, and in red blood cells of Inuits,45 and
Hg-bound SelP has been identified in the plasma of Inuits 46 and miners exposed to mercury47) and
in the liver, kidneys, and muscle of the waterbird Clark’s grebe (Aechmophorus clarkii).43 The
demethylated mercury atoms are coordinated to four selenocysteine residues (Hg(Sec)4 complex) in
SelP.43 The coordination number of Hg with selenoneine is unknown, however, it is probably also
four because this coordination is easy to obtain at Se/Hg = 4 with synthetic derivatives analogue to
selenoneine,48 whereas it requires a high excess of sulfur with thiolate ligands.49, 50 This difference is
explained, on one hand by the greater selenophilicity than thiophilicity of Hg,15 and on the other hand
by the better nucleophilicity and higher acidity of the selenolate anion (pKa(-SeH) = 5.4) than the
thiolate anion (pKa(-SH) = 8.2).51 Thus, tetrahedral bonding with four selenium atoms appears to be
the preferred coordination of Hg in organic molecules. The tetraselenolate complex having a Hg:Se
molar ratio of 1:4 compared to 1:1 for HgSe, all the Hg can be bound to Se at a Hg:Se ratio much
lower than 1:1. Therefore, overlooking the Hg(Sec)4 species leads to overestimate the amount of
bioavailable Se and in turn to underestimate the toxicity of mercury.
Here, we report the chemical forms of mercury in the feathers, blood, liver, kidneys, muscle, and
brain of giant petrels (Macronectes spp), as determined using Hg L3-edge high energy-resolution X-
ray absorption near edge structure (HR-XANES)43, 52-62 and extended X-ray absorption fine structure
(EXAFS) spectroscopy,63 high resolution transmission electron microscopy (HRTEM), and high-
angle annular dark-field scanning transmission electron microscopy (STEM-HAADF) coupled to
elemental mapping from energy dispersive X-ray spectrometry (EDX). Coupled with chemical
analyses, the quantitative speciation data allowed calculating the concentration of bioavailable Se
([Se]bio), the molar fraction of bioavailable Se to total Se ([Se]bio:[Se]tot), and the effective (or
biological) fraction of Se bound to Hg ((Hg:Se)eff), and to discuss the toxicity of mercury from a
speciation perspective.
MATERIALS AND METHODS
4
A detailed description of the bird samples, experimental methods, and data analysis is given in the
Supporting Information (SI).
Samples. Eight dead giant petrels were opportunistically collected in French Southern and
Antarctic Lands (Table S1). Individuals were stored at −20 °C until dissection and the tissues
lyophilized for chemical analysis and spectroscopic measurement. Freeze-drying a frozen tissue does
not change the speciation of the metal.56, 64 Age and breeding status of the birds are not known but all
are adult males. Because females can eliminate part of their Hg load trough egg production, we
selected males to avoid the effect of maternal transfer on Hg metabolism. We studied the Hg
speciation in the feathers, liver, kidneys, and muscle from five petrels (P1-P5), in the blood from
three out of the five (P1-P3), and in the brain from three other birds (P6-P8) (Table S2). The 26 tissues
were analyzed by HR-XANES, and two liver, two kidney, and three muscle tissues by EXAFS.
Hg and Se analyses. Total mercury was quantified with an AMA-254 mercury analyzer (Altec,
Prague; limit of detection, LOD, 0.005 µg/g dw; aliquot mass: 0.5–5 mg dw) and total selenium with
a Thermo Fisher Scientific X Series 2 ICP-MS (LOD 0.08 µg/g dw; aliquots mass: 20–300 mg dw).
HR-XANES and EXAFS spectroscopy. Mercury L3-edge HR-XANES and EXAFS spectra were
measured at 10-15 K with high-reflectivity analyzer crystals65 on beamline ID26 at the European
Synchrotron Radiation Facility (ESRF). HR-XANES data were analyzed by principal component
analysis (PCA66) and target transformation67, 68 against a large database of spectra described
previously.43, 52, 54-56, 59 All reference spectra were considered as a basis for identification of the Hg
species, but only diagnostic spectra are discussed herein. The molar proportions (mol%) of the Hg
species in the petrel tissues was obtained by linear combination fitting analysis. The accuracy of the
amount of a fitted component was estimated to be equal to the variation of its value when the fit
residual (NSS) was increased by 20%. NSS is the normalized sum-squared difference between two
spectra expressed as Σ[(yexp-yfit)2]/Σ(yexp
2).
Electron microscopy. Scanning transmission electron microscopy (STEM) images were acquired
on Nant'Themis, a Themis Z G3 Cs-probe corrected microscope (Thermo Fisher Scientific) operated
at 80 kV (or 300 kV in cryo mode) and equipped with a high-angle annular dark field (HAADF)
detector (resolution ~1.0 Å and 80 kV). Elemental maps were acquired in STEM mode using a Super-
XTM emission X-ray spectrometer consisting of four windowless silicon-drift detectors (SDDs)
providing a large collection solid angle of 0.7 srad. Additional high-resolution TEM (HRTEM)
images and selected-area electron diffraction (SAED) patterns were acquired at 300 kV on a GATAN
5
OneView camera (resolution ~1.8 Å). The liver, kidney, and muscle tissues of individual P4 (Tables
S1 and S2) were dispersed in ethanol and deposited on a lacey-C copper grid.
RESULTS
Mercury and selenium concentrations. The amounts of Hg and Se are extremely variable
between and within birds. The dry weight concentrations of Hg range from 170 to 1499 μg/g in liver
(n = 7), from 9.8 to 414 μg/g in kidneys (n = 8), from 2.9 to 88.7 μg/g in muscle (n = 8), from 4.8 to
26.2 μg/g in feathers (n = 8), from 1.3 to 23.9 μg/g in blood (n = 8), and from 1.6 to 13.2 μg/g in brain
(n = 3) (Table S2). The majority of the Hg concentrations are close to, and sometimes above, the
highest values reported to date in seabird tissues,6-8, 69, 70 but remain below values reported for marine
mammals.71, 72 The corresponding ranges of Se concentrations are 109-1101 μg/g dw in liver (n = 7),
55-363 μg/g in kidneys (n = 8), 10.1-72.7 μg/g in muscle (n = 8), 5.9-24.2 μg/g in feathers (n = 8),
43.4-118 μg/g in blood (n = 8), and 26.2-47.0 μg/g in brain (n = 3). The liver always contains the
highest levels of Hg and Se, as a result of its detoxification role and the toxicological Hg-Se
antagonism.19 The concentrations of Hg and Se in non-hepatic tissues relative to their concentrations
in liver decrease in the following order: for Hg, liver (100%) > kidneys (14.5 ± 8.6%) > muscle (5.9
± 3.5%) > feathers (3.8 ± 1.9%) > brain (2.3 ± 1.4%) > blood (1.9 ± 2.0%), and for Se, liver (100%)
> kidneys (65.5 ± 39.5%) > blood (42.6 ± 31.2%) > brain (31.3 ± 9.5%) > muscle (15.3 ± 8.4%) >
feathers (6.1 ± 7.0%) (Figure 1). The blood selenium level is particularly high, between 0.5 and 1.2
μmol/g, which is 30 to 90 times more than Hg on a molar basis. The high blood selenium is explained
by the Se-rich diet and by the metabolic function of the liver, which sequesters Hg and secretes SelP
into the plasma to supply Se to endocrine tissues to synthesize selenoproteins they need for their
metabolism.38 33, 38, 73-76
The relationship between the molar concentrations of Se and Hg in petrel tissues is shown in Figure
2a. The Se:Hg molar ratio is always higher than 1:1, suggesting that Se is always in excess relative
to Hg, therefore not limiting for the detoxification of MeHg. For the liver, kidneys, and muscle, there
is a trend of higher Se content with more Hg, which fuels the idea that MeHg is at least partly
demethylated by selenolate groups48, 77 in these tissues. The Se and Hg values are somewhat aligned.
A regression analysis shows that the Hg and Se concentrations are correlated in the liver (R2 = 0.99).
A high Se:Hg ratio is distinctive of the blood (52.2 ± 22.5) and the brain (29.9 ± 18.1). A high Se:Hg
ratio in the blood is consistent with the transport and elimination from the circulatory system of the
MeHg species, which has no selenium, and the delivery of essential Se to peripheral tissues. A high
6
Se:Hg ratio in the brain is intriguing because, while the Hg content is lower than in the other organs
(5.9 ± 6.4 μg/g), yet the proportion of brain Se to whole-body Se is relatively high (Table S2). The
brain has a smaller selenoproteome than the liver and kidneys, being normally similar to that of the
muscle.38 The Se level in brain, as expressed as a percentage of the liver level, is twice that of the Se
level in muscle (31.3 ± 9.5 μg/g vs. 15.3 ± 8.4 μg/g; Figure 1b). Actually, brain neurons require a
reliable supply of selenium for their functioning, as selenium deficiency causes neurological
impairment.33, 38, 73-76 SelP is delivered to neurons (and endocrine cells) upon binding to the cell-
surface apolipoprotein E Receptor-2 (ApoER2).78, 79
Chemical forms of Hg. The HR-XANES spectra from the 26 petrel tissues and the EXAFS spectra
from eight of them are represented in Figures S1 and S2a. All HR-XANES spectra of feathers were
identified as being from a MeHg-cysteinate complex (MeHgCys) and two liver spectra (P1, P2) from
HgSe. All other spectra have variable shapes distinct from any standard, implying that they are a
weighted sum of single-species spectra. There is a large inter- and intra-individual variability in tissue
speciation, represented in Figures 3a with the brain spectra from individuals P6, P7, and P8, and in
Figure 3b with the liver, kidneys, and muscle spectra from a same individual (P2). The compositional
variability is also observed on EXAFS spectra (Figure S2b). Investigating all 26 HR-XANES spectra
using PCA, we found that four principal components (species) were required to account for the
variance in the data set (Figure S3). The single-species spectra needed to describe quantitatively the
HR-XANES data set, as identified by target transformation, are shown in Figure 3c-f together with
the structure of the Hg species.
Besides MeHgCys and HgSe, two inorganic Hg complexes were identified, a dithiolate (Hg(SR)2)
and a tetraselenolate (Hg(SeR)4) complex. Since the two complexes occur in biological matter, the
thiolate group is from a cysteine residue (Hg(Cys)2) and the selenolate group from a selenocysteine
group (Hg(Sec)4). The spectrum from the dicysteinate complex Hg(GSH)2 at physiological pH55
provided the best reconstruction by target transformation for the first complex (NSS = 1.5 10-4), and
the Hg(Sec)4 spectrum43 provided the best reconstruction for the second complex (NSS = 4.8 10-5).
Consideration of the Hg(Sec)4 complex is also needed to explain the variability of the EXAFS spectra
(Figure S2b). In Hg(GSH)2, Hg is coordinated primarily to two cysteine residues and secondarily to
two oxygen atoms in a seesaw geometry,59 also denominated disphenoidal (Figure 3). The
Hg[(Cys)2+(N/O)1-2] coordination has been identified previously in the Clark’s grebe (in addition to
Hg(Sec)4), fish, earthworm, human hair, and bacteria.43, 54, 59, 60, 62
7
The molar proportions, or fractional amount f, of the Hg species in the petrel tissues, as obtained
by least-squares fitting of the tissue spectra with linear combinations of the four single-species spectra
identified by PCA, are given in Table S2. The tissue proportion of each Hg species, averaged over
the petrel individuals, is represented in the diagram of Figure 4. The weight concentrations of the Hg
species are obtained by multiplying f by the Hg concentration. The main results from the HR-XANES
and EXAFS analysis are the following:
• MeHgCys occurs in all tissues, except the liver (detection limit 3% of total Hg). Its proportion
decreases in the following order: feathers (100%, n = 5) > blood (83 ± 20%, n = 3) > brain
(45 ± 35%, n = 3) > kidneys (4 ± 5%, n = 5) = muscle (4 ± 5%, n = 5).
• HgSe and Hg(Sec)4 occur in all tissues, except the feathers and blood (detection limit 9% of
total Hg). The proportion of HgSe decreases in the order: liver (95 ± 5%, n = 5) > kidneys (61
± 8%, n = 5) > brain (38 ± 32%, n = 3) = muscle (35 ± 15%, n = 5). The proportion of Hg(Sec)4
decreases in the order: muscle (61 ± 13%, n = 5) > kidneys (35 ± 10%, n = 5) > brain (16 ±
7%, n = 3) > liver (5 ± 2%, n = 5).
• MeHgCys and HgSe never co-occur without Hg(Sec)4.
• Hg(Cys)2 (modeled with Hg(GSH)2) occurs in the blood of petrels P3 (12 ± 9%) and P5 (39
± 8%) together with MeHgCys, but not in the blood of petrel P1. The low abundance of
Hg(Cys)2 in the data set (2 occurrences out of 26), and relatively low signal-to-noise ratio of
its HR-XANES spectrum (Figure S1), explain the poorer reconstruction of the Hg(GSH)2
proxy by target transformation: NSS = 1.5 10-4, compared to NSS = 1.4 10-6 for HgSe, NSS =
4.8 10-5 for Hg(Sec)4, and NSS = 3.7 10-5 for MeHgCys (Figure 3c-f).
• The HgSe grains do not contain sulfur, therefore do not form a Hg(SexS1-x) solid-solution, in
contrast to previous hypotheses22’80 (Figure S4, Table S3).
Imaging of HgSe nanoparticles. Sparsely distributed electron-dense aggregates of 5 nm up to
100 nm HgSe nanocrystals were imaged, for the first time in seabird tissues, using electron
microscopy (Figures 5 and S5-S13). Although large variations in crystal size were observed in each
tissue, overall it decreased with the Hg concentration from about 40-100 nm in liver (1499 μg Hg/g
dw), to 5-20 nm in kidneys (414 μg Hg/g), to ~3 nm in muscle (88.7 μg Hg/g). Few aggregates of ~3
nm and larger crystals also were observed in the liver. Summation of the EDX spectra from all
individual pixels of the large crystal shown in Figure 5a gave a Hg:Se atomic ratio of ~1. The sulfur
intensity profile did not increase over the grain compared to the surrounding tissue, negating a
Hg(Se,S) solid solution (Figure S5). The electron diffraction (SAED) pattern and interplanar
8
distances obtained by fast-Fourier transform (FFT) matched the Fd3m cubic space group and a = 6.08
Å lattice dimension of HgSe81, which is isostructural to β-HgS (Figure 5a). Crystallographic data and
nanochemical analyses performed on many nanocrystals of varying size in the three tissues were all
consistent with HgSe. The smallest crystals having lattice planes which could be imaged had a
dimension of 4-5 nm. However, STEM imaging of kidney and muscle aggregates show that they
contain vanishingly small particles, down to about 1 nm, and perhaps below. The EDX signal from
the small particles was too low to measure their Hg:Se ratio.
DISCUSSION
This study provides the first evidence in seabird tissues of the Hg(Sec)4 species identified recently in
Clark’s grebe and freshwater fish.43 This finding warrants further discussion on its significance for
the Hg-Se antagonism, the Se metabolism, and the demethylation of methylmercury as HgSe.
The Hg-Se antagonism. When Hg is bound in equimolar stoichiometry with Se as in HgSe, the
Hg:Se atomic ratio derived from chemical analysis ((Hg:Se)chem) represents the fraction of Se, to total
Se, bound to Hg. However, when Hg is complexed to four selenolate ligands, forming a Hg(Sec)4
complex, the effective fraction of Se bound to Hg is higher than the chemical fraction ((Hg:Se)eff >
(Hg:Se)chem). The difference between the two molar ratios depends on the fractional amount, or
proportion, of Hg(Sec)4 and HgSe in the sample, noted fHg(Sec)4 and fHgSe. The effective Hg:Se ratio
can be written:
(Hg:Se)eff = (Hg:Se)chem x (fHgSe + 4 x fHg(Sec)4)
For example, the liver of petrel P3 has (Hg:Se)chem = 0.54, fHgSe = 0.91, fHg(Sec)4 = 0.09, which gives
(Hg:Se)eff = 0.69. The concentration of bioavailable Se is [Se]bio = (1 - (Hg:Se)eff) x [Se]tot. In this
example, [Se]tot = 213 μg/g and [Se]bio = 66 μg/g (Table S2). Omitting the 1:4 stoichiometric ratio of
Hg(Sec)4 gives an apparent concentration of appbio[Se] = (1 - 0.54 ) x 213 = 98 μg/g, 48% higher than
the actual concentration. Similarly, the kidneys of petrel P4 have [Se]tot = 363 μg/g, appbio[Se] = (1 –
0.45) x 363 = 200 μg/g, and [Se]bio = 58 μg/g. The apparent excess of Se obtained by omitting the
Hg(Sec)4 complex is more than three times higher than the actual value. Overestimating [Se]bio is not
a key issue in liver and kidneys as Se concentration is not limiting in these tissues. In contrast,
overestimating [Se]bio leads to underestimate the toxicological risk, as discussed below for muscle.
The relationships between (Hg:Se)chem and [Hg]tot and between (Hg:Se)eff and [Hg]tot in the 26
tissues are shown in Figures 6a and 6b, and the fraction of bioavailable Se to total Se ([Se]bio/[Se]tot)
in Figure 6c. Molar Hg:Se ratios vary greatly among all tissues and between individuals as shown
9
previously in Figure 2a. Incorporating fHg(Sec)4 in the calculation of Hg:Se has the most effect on the
muscular tissue. In three out of five birds (P2, P3, P4), muscular (Hg:Se)eff is close to 1:1 (Figure
6b), the threshold from which toxic effects may emerge.37 The accuracy on ratios is highly variable,
however, as a result of the propagation of uncertainties from chemical analyses and spectroscopic
values of fHgSe and fHg(Sec)4. Notwithstanding these cumulative errors, it appears clearly that the muscle
is the most affected of all tissues by Hg contamination. For example, the muscle of petrel P2 has a
few μg/g of bioavailable Se at most, for an apparent concentration of appbio[Se] = (1 – 0.33) x 50.8 = 34
μg/g (Table S2).
Figure 2a also can be corrected for fHg(Sec)4 like Figure 6a. In this case [Se]tot on the y-scale needs
to be replaced by the effective Se concentration, which is
[Se]eff = (1/(Hg:Se)eff) x [Hg]tot = (Se:Hg)eff x [Hg]tot
with [Hg]tot and [Se]eff in mol/g. In this representation, the muscle tissues of the three birds that are
depleted in bioavailable Se (P2, P3, P4) are aligned on the 1:1 theoretical line (Figure 2b). Blood
results cannot be represented on this graph because Hg is not bound to Se, but is in methylated and
dicysteinate form only (MeHgCys + Hg(Cys)2). For this tissue, the (Se:Hg)eff ratio, which represents
the excess of Se, is infinite.
Se metabolism. Selenium is an essential element which is distributed by the liver to non-hepatic
tissues to maintain Se homeostasis.33, 38 Selenium content is well-regulated in laboratory animals fed
with dietary Se intakes. In mice, the regulated selenium pool is about 200 ng/g wet weight (~1 μg/g
dw) in muscle and 1500 ng/g wet weight (~6 μg/g dw) in liver.82 As intake increases, excess Se is
excreted predominantly via urine. The bioavailable concentration of Se in the muscle of the three
most depleted birds (P2, P3, P4) is on the order of 1 ± 10 μg/g dw (i.e., 0.1-10 μg/g). Although
muscular tissues are most impacted by Hg contamination, a Hg-induced Se-deficiency is unlikely
because the liver contains enough bioavailable Se for replenishment (Figure S14a). The liver of the
two most contaminated birds (P2, P4), containing 804 μg/g and 1499 μg Hg/g dw, respectively, are
concomitantly in tremendous excess of [Se]bio (Figure S14b). The [Se]bio:[Se]tot ratio is approximately
constant ranging between 0.28 and 0.49 (Figure 14c).
Source of the Hg-dicysteinate species. The Hg(Cys)2 species was detected in the blood of
individuals P3 and P5, not P1, and co-occurs with MeHgCys. The source of MeHg is clearly
exogenous since it is the main form of Hg in the preys of petrel, such as fish and squids,83-86 and is
almost totally (90−95%) absorbed in the gastrointestinal tract.87, 88 The Hg(Cys)2 species is probably
exogenous too, otherwise Hg would be bonded to Se, not to S, as no MeHgCys to Hg(Cys)2
10
demethylation pathway is known in the body to date. Gastrointestinal absorption of Hg(Cys)2,
although inefficient (7-15%), occurs in the lumen of the intestine.89, 90 Hg(Cys)2 can be produced by
demethylating intestinal microflora or taken up through the diet. Giant petrels eat fish and
cephalopods when they are at sea and they are very aggressive predators and scavengers of mammal’s
and bird’s carcasses when on land. The high variation of the %Hg(Cys)2 to %MeHgCys ratio in blood
among the three birds (P1: 0% / 100%; P3: 12% / 88%; P5: 39% / 61%) argues for a dietary source
that differed at the time of sampling. Thus, Hg speciation in blood informs about recent intakes. The
lack of Hg(Cys)2 in other tissues can be explained by a Se for S ligand exchange and by its elimination
from the bloodstream in urine.91
Demethylation of methylmercury. It has been suggested that the HgSe demethylation reaction
is a concentration-sensitive process in the liver of dolphins and waterbirds, starting when a mercury
concentration threshold is reached.92, 93 In waterbirds the threshold value is approximately 8.5 μg
Hg/g dw. Here, there is no evidence for a demethylation threshold because the liver concentrations
are at least 20 times higher than the waterbird threshold ([Hg]tot >170 μg/g). Because the vast majority
of Hg is inorganic and %HgSe ≥ 91 ± 6%, quantities of MeHg are very low (Table S2). In contrast to
liver, there is some supporting evidence for a HgSe threshold in brain, as its Hg concentration ranges
from 1.6 μg/g (petrel P7) to 13.2 μg /g (P8). The response of %MeHg to changing brain [Hg]tot is
strongly curvilinear (Figure S14c). %MeHg declines abruptly from 83 ± 5% at [Hg]tot = 1.6 μg/g, to
40 ± 6% at [Hg]tot = 2.8 μg/g, then less rapidly to 13 ± 5% at [Hg]tot = 13.2 μg/g. In contrast, %HgSe
and %(HgSe+Hg(Sec)4) follow a power-law with [Hg]tot. HgSe amounts to 8 ± 8% at [Hg]tot = 1.6
μg/g, 36 ± 13% at [Hg]tot = 2.8 μg/g, and 71 ± 8% at [Hg]tot = 13.2 μg/g (Table S2, Figure S14d).
This suggests that (1) brain MeHg accumulates to a threshold concentration of about 1-2 μg/g above
which HgSe is formed, and (2) HgSe nucleation is initiated by the Hg(Sec)4 complex, in agreement
with electron microscopy observations and previous results on the Clark’s grebe.43 The fact that
MeHgCys and HgSe never co-exist in any tissue without Hg(Sec)4 is another supporting line of
evidence for Hg(Sec)4 being an intermediate species of the demethylation reaction. Hg(Sec)4 and
HgSe observed in kidneys, muscle, and brain are formed in situ from MeHgCys and SelP delivered
by the bloodstream, and also from SelP directly produced in peripheral tissues.76, 94
The low demethylation threshold value observed in the brain, together with a [Se]tot/[Hg]tot molar
ratio of 30 ± 18 (Figure 2a), many times higher than in the other tissues, suggest that the brain is the
first organ protected from injury. This finding agrees with Se-deficiency experiments on animal
models, which show that neurons take up Se via the ApoER2 receptor at the expense of kidneys and
11
muscle to prevent injury.38, 73, 76 Demethylation of MeHg in HgSe has been observed previously in
the brain of dolphins26 and humans.95
Molecular structure of the Hg-tetraselenolate complex. We showed recently that the Hg(Sec)4
complex is bound to selenoprotein P in the Clark’s grebe.43 SelP is present in all vertebrates and
harbors at least ten Sec residues, compared to at most two in other selenoproteins.96 Birds have one
Sec in the α-domain and twelve in the β-domain, seven of which being grouped in the carboxyl-
terminus region of the protein.43 The last four Sec residues of the protein tail are arranged in the
highly conserved amino acid motif XUXUX6UXUX (single-letter amino acid code, where U is Sec
and X can be any amino acid). With its four Sec residues, this region is the most likely Hg-binding
site. The tertiary structure of the predicted Hg-binding site in giant petrel was modeled by iterative
threading assembly refinement (I-TASSER97, 98) from the amino acid sequence of the northern fulmar
(Fulmarus glacialis). I-TASSER matches with Monte Carlo simulations the structure of 3D models
with known protein structures in the Protein Data Bank (PDB). SelP of giant petrel has not been
sequenced, but it should be close to that of northern fulmar, as it is phylogenetically close (both
belong to the fulmarine group within the same Procellariidae family). The two UXU motifs are
predicted to face each other on each side of a β-turn, forming a four-coordinate metal-binding site
(Figure 7).
The association in Hg-Se aggregates of <4 nm nanoparticles with ≥ 4 nm HgSe crystals (Figures
S10, S11, and S13) is suggestive of a continuum of Hg cluster sizes from Hgx(Se,Sec)y multinuclear
complexes to HgSe nanocrystals. We hypothesize that the flexible Sec-rich β-domain of SelP forms
a multinuclear binding pocket via protein folding and mediates the biomineralization of HgSe via
self-assembling. This mechanism is common in the biosynthesis of metalloclusters,99-101 and is
reminiscent of metal clustering in metallothioneins.56, 102, 103 Then, the Hg(Sec)4 complex would act
as an external ligand to the nascent HgSe nanoparticles, decreasing their Hg:Se molar ratio below 1.
Supporting this possibility is the deviation of the equimolar ratio observed near HgSe grains in the
liver of dolphins by Gajdosechova and coworkers.26 We considered all along this study that the
Hg(Sec)4 complex is mononuclear. If the β-domains of SelP host multinuclear Hgx(Se,Sec)y
complexes, then they would be too disordered to be identified as a distinct species by HR-XANES.
A caveat in assigning all tetraselenolate complexes to Hg(Sec)4 is an overestimation of (Hg:Se)eff,
because y < 4x in Hgx(Se,Sec)y.43 We also considered in this study that Hg is the only potent
electrophile capable of binding Se. However, there are additional elements of environmental concern
which can also bind it and contribute to loss of Se that is biologically available for selenoenzyme
12
synthesis. Organic electrophiles33, 104 bioaccumulated in polluted ecosystems may contribute to
impair Se bioavailability.
REFERENCES
(1) Gonzalez-Solis, J.; Sanpera, C.; Ruiz, X., Metals and selenium as bioindicators of geographic
and trophic segregation in giant petrels Macronectes spp. Mar. Ecol. Prog. Ser. 2002, 244, 257-264.
(2) Carvalho, P. C.; Bugoni, L.; McGill, R. A. R.; Bianchini, A., Metal and selenium
concentrations in blood and feathers of petrels of the genus procellaria. Environ. Toxicol. Chem. 2013,
32, 1641-1648.
(3) Bustamante, P.; Carravieri, A.; Goutte, A.; Barbraud, C.; Delord, K.; Chastel, O.;
Weimerskirch, H.; Cherel, Y., High feather mercury concentrations in the wandering albatross are
related to sex, breeding status and trophic ecology with no demographic consequences. Environ. Res.
2016, 144, 1-10.
(4) Carravieri, A.; Cherel, Y.; Brault-Favrou, M.; Churlaud, C.; Peluhet, L.; Labadie, P.;
Budzinski, H.; Chastel, O.; Bustamante, P., From Antarctica to the subtropics: Contrasted
geographical concentrations of selenium, mercury, and persistent organic pollutants in skua chicks
(Catharacta spp.). Environ. Poll. 2017, 228, 464-473.
(5) Sebastiano, M.; Bustamante, P.; Eulaers, I.; Malarvannan, G.; Mendez-Fernandez, P.;
Churlaud, C.; Blevin, P.; Hauselmann, A.; Covaci, A.; Eens, M.; Costantini, D.; Chastel, O., Trophic
ecology drives contaminant concentrations within a tropical seabird community. Environ. Poll. 2017,
227, 183-193.
(6) Albert, C.; Renedo, M.; Bustamante, P.; Fort, J., Using blood and feathers to investigate large-
scale Hg contamination in Arctic seabirds: A review. Environ. Res. 2019, 177, n° 108588.
(7) Gilmour, M. E.; Lavers, J. L.; Lamborg, C.; Chastel, O.; Kania, S. A.; Shaffer, S. A., Mercury
as an indicator of foraging ecology but not the breeding hormone prolactin in seabirds. Ecol. Indic.
2019, 103, 248-259.
(8) Renedo, M.; Pedrero, Z.; Amouroux, D.; Cherel, Y.; Bustamante, P., Mercury isotopes of key
tissues document mercury metabolic processes in seabirds. Chemosphere 2021, 263, n° 127777.
(9) Norheim, G., Levels and interactions of heavy-metals in sea birds from Svalbard and the
Antarctic. Environ. Poll. 1987, 47, 83-94.
(10) Kim, E. Y.; Goto, R.; Tanabe, S.; Tanaka, H.; Tatsukawa, R., Distribution of 14 elements in
tissues and organs of oceanic seabirds. Arch. Environ. Con. Tox. 1998, 35, 638-645.
13
(11) Ikemoto, T.; Kunito, T.; Tanaka, H.; Baba, N.; Miyazaki, N.; Tanabe, S., Detoxification
mechanism of heavy metals in marine mammals and seabirds: Interaction of selenium with mercury,
silver, copper, zinc, and cadmium in liver. Arch. Environ. Contam. Toxicol. 2004, 47, (3), 402-413.
(12) Burger, J.; Jehl, J. R.; Gochfeld, M., Selenium:mercury molar ratio in eared grebes (Podiceps
nigricollis) as a possible biomarker of exposure. Ecol. Indic. 2013, 34, 60-68.
(13) Cipro, C. V. Z.; Cherel, Y.; Caurant, F.; Miramand, P.; Mendez-Fernandez, P.; Bustamante,
P., Trace elements in tissues of white-chinned petrels (Procellaria aequinoctialis) from Kerguelen
waters, Southern Indian Ocean. Polar Biol. 2014, 37, 763-771.
(14) Moura, J. F.; Tavares, D. C.; Lemos, L. S.; Acevedo-Trejos, E.; Saint'Pierre, T. D.; Siciliano,
S.; Merico, A., Interspecific variation of essential and non-essential trace elements in sympatric
seabirds. Environ. Poll. 2018, 242, 470-479.
(15) Melnick, J. G.; Yurkerwich, K.; Parkin, G., On the chalcogenophilicity of mercury: Evidence
for a strong Hg-Se bond in [TmBut]HgSePh and its relevance to the toxicity of mercury. J. Am. Chem.
Soc. 2010, 132, 647-655.
(16) Ohi, G.; Nishigaki, S.; Seki, H.; Tamura, Y.; Maki, T.; Konno, H.; Ochiai, S.; Yamada, H.;
Shimamura, Y.; Mizoguchi, I.; Yagyu, H., Efficacy of selenium in tuna and selenite in modufying
methylmercury intoxication. Environ. Res. 1976, 12, (1), 49-58.
(17) Dauplais, M.; Lazard, M.; Blanquet, S.; Plateau, P., Neutralization by metal ions of the
toxicity of sodium selenide. Plos One 2013, 8, n° e54353.
(18) Gailer, J.; George, G. N.; Pickering, I. J.; Madden, S.; Prince, R. C.; Eileen Y. Yu; Denton,
M. B.; Younis, H. S.; Aposhian, H. V., Structural basis of the antagonism between inorganic mercury
and selenium in mammals. Chem. Res. Toxicol. 2000, 13, 1135-1142.
(19) Cuvin-Aralar, M. L. A.; Furness, R. W., Mercury and Selenium interaction: a review. Ecotox.
Environ. Saf. 1991, 21, 348-364.
(20) Yang, D. Y.; Chen, Y. W.; Gunn, J. M.; Belzile, N., Selenium and mercury in organisms:
Interactions and mechanisms. Environ. Rev. 2008, 16, 71-92.
(21) Li, M. L.; Juang, C. A.; Ewald, J. D.; Yin, R. S.; Mikkelsen, B.; Krabbenhoft, D. P.; Balcom,
P. H.; Dassuncao, C.; Sunderland, E. M., Selenium and stable mercury isotopes provide new insights
into mercury toxicokinetics in pilot whales. Sci. Tot. Environ. 2020, 710, n° 136325.
(22) Arai, T.; Ikemoto, T.; Hokura, A.; Terada, Y.; Kunito, T.; Tanabe, S.; Nakai, I., Chemical
forms of mercury and cadmium accumulated in marine mammals and seabirds as determined by
XAFS analysis. Environ. Sci. Technol. 2004, 38, 6468-6474.
14
(23) Nakazawa, E.; Ikemoto, T.; Hokura, A.; Terada, Y.; Kunito, T.; Tanabe, S.; Nakai, I., The
presence of mercury selenide in various tissues of the striped dolphin: evidence from -XRF-XRD
and XAFS analyses. Metallomics 2011, 3, 719-725.
(24) Lailson-Brito, J.; Cruz, R.; Dorneles, P. R.; Andrade, L.; Azevedo, A. D.; Fragoso, A. B.;
Vidal, L. G.; Costa, M. B.; Bisi, T. L.; Almeida, R.; Carvalho, D. P.; Bastos, W. R.; Malm, O.,
Mercury-selenium relationships in liver of Guiana dolphin: The possible role of Kupffer cells in the
detoxification process by tiemannite formation. Plos One 2012, 7, n° e42162.
(25) Sakamoto, M.; Itai, T.; Yasutake, A.; Iwasaki, T.; Yasunaga, G.; Fujise, Y.; Nakamura, M.;
Murata, K.; Chan, H. M.; Domingo, J. L.; Marumoto, M., Mercury speciation and selenium in
toothed-whale muscles. Environ. Res. 2015, 143, 55-61.
(26) Gajdosechova, Z.; Lawan, M. M.; Urgast, D. S.; Raab, A.; Scheckel, K. G.; Lombi, E.;
Kopittke, P. M.; Loeschner, K.; Larsen, E. H.; Woods, G.; Brownlow, A.; Read, F. L.; Feldmann, J.;
Krupp, E. M., In vivo formation of natural HgSe nanoparticles in the liver and brain of pilot whales.
Sci. Rep. 2016, 6, 34361.
(27) Huggins, F. E.; Raverty, S. A.; Nielsen, O. S.; Sharp, N. E.; Robertson, J. D.; Ralston, N. V.
C., An XAFS investigation of mercury and selenium in Beluga whale tissues. Environ. Bioindic.
2009, 4, 291-302.
(28) Khan, M. A. K.; Wang, F. Y., Chemical demethylation of methylmercury by selenoamino
acids. Chem. Res. Toxicol. 2010, 23, 1202-1206.
(29) Ralston, N. V. C.; Blackwell, J. L.; Raymond, L. J., Importance of molar ratios in selenium-
dependent protection against methylmercury toxicity. Biol. Trace Elem. Res. 2007, 119, 255-268.
(30) Peterson, S. A.; Ralston, N. V. C.; Peck, D. V.; Van Sickle, J.; Robertson, J. D.; Spate, V. L.;
Morris, J. S., How might selenium moderate the toxic effects of mercury in stream fish of the western
US? Environ. Sci. Technol. 2009, 43, 3919-3925.
(31) Sormo, E. G.; Ciesielski, T. M.; Overjordet, I. B.; Lierhagen, S.; Eggen, G. S.; Berg, T.;
Jenssen, B. M., Selenium moderates mercury toxicity in free-ranging freshwater fish. Environ. Sci.
Technol. 2011, 45, 6561-6566.
(32) Kehrig, H. A.; Seixas, T. G.; Di Beneditto, A. P. M.; Malm, O., Selenium and mercury in
widely consumed seafood from South Atlantic Ocean. Ecotoxicol. Environ. Safety 2013, 93, 156-162.
(33) Ralston, N. V. C.; Raymond, L. J., Mercury's neurotoxicity is characterized by its disruption
of selenium biochemistry. BBA-Gen. Sub. 2018, 1862, (11), 2405-2416.
15
(34) Yu, X. P.; Wang, M. X.; Nan, X. J.; Guo, Y. F.; Deng, T. L., Species and correlations between
selenium and mercury in fishpond ecosystems. Water Environ. Res. 2019, 91, 292-299.
(35) Ralston, N. V. C.; Ralston, C. R.; Blackwell, J. L.; Raymond, L. J., Dietary and tissue selenium
in relation to methylmercury toxicity. Neurotoxicol. 2008, 29, 802-811.
(36) Calatayud, M.; Devesa, V.; Virseda, J. R.; Barbera, R.; Montoro, R.; Velez, D., Mercury and
selenium in fish and shellfish: Occurrence, bioaccessibility and uptake by Caco-2 cells. Food Chem.
Toxicol. 2012, 50, (8), 2696-2702.
(37) Mulder, P. J.; Lie, E.; Eggen, G. S.; Ciesielski, T. M.; Berg, T.; Skaare, J. U.; Jenssen, B. M.;
Sormo, E. G., Mercury in molar excess of selenium interferes with thyroid hormone function in free-
ranging freshwater fish. Environ. Sci. Technol. 2012, 46, 9027-9037.
(38) Burk, R. F.; Hill, K. E., Regulation of selenium metabolism and transport. In Annu. Rev. Nutr.,
Bowman, B. A.; Stover, P. J., Eds. 2015; Vol. 35, pp 109-134.
(39) Falnoga, I.; Tusek-Znidaric, M.; Stegnar, P., The influence of long-term mercury exposure on
selenium availability in tissues: An evaluation of data. Biometals 2006, 19, 283-294.
(40) Cusack, L. K.; Eagles-Smith, C.; Harding, A. K.; Kile, M.; Stone, D., Selenium: mercury
molar ratios in freshwater fish in the Columbia River basin: Potential applications for specific fish
consumption advisories. Biol. Trace Elem. Res. 2017, 178, 136-146.
(41) Zhang, H.; Feng, X. B.; Chan, H. M.; Larssen, T., New insights into traditional health risk
assessments of mercury exposure: Implications of selenium. Environ. Sci. Technol. 2014, 48, 1206-
1212.
(42) Yamashita, M.; Yamashita, Y.; Suzuki, T.; Kani, Y.; Mizusawa, N.; Imamura, S.; Takemoto,
K.; Hara, T.; Hossain, M. A.; Yabu, T.; Touhata, K., Selenoneine, a novel selenium-containing
compound, mediates detoxification mechanisms against methylmercury. Accumulation and toxicity
in zebrafish embryo. Mar. Biotechnol. 2013, 15, 559-570.
(43) Manceau, A.; Bourdineaud, J. P.; Oliveira, R. B.; Sarrazin, S. L. F.; Krabbenhoft, D. P.;
Eagles-Smith, C. A.; Ackerman, J. T.; Stewart, A. R.; Ward-Deitrich, C.; Busto, M. E. D.; Goenaga-
Infante, H.; Wack, A.; Retegan, M.; Detlefs, B.; Glatzel, P.; Bustamante, P.; Nagy, K. L.; Poulin, B.
A., Demethylation of methylmercury in bird, fish, and earthworm. Environ. Sci. Technol. 2020,
Submitted.
(44) Yamashita, Y.; Yamashita, M., Identification of a novel selenium-containing compound,
selenoneine, as the predominant chemical form of organic selenium in the blood of bluefin tuna. J.
Biol. Chem. 2010, 285, 18134-18138.
16
(45) Achouba, A.; Dumas, P.; Ouellet, N.; Little, M.; Lemire, M.; Ayotte, P., Selenoneine is a
major selenium species in beluga skin and red blood cells of Inuit from Nunavik. Chemosphere 2019,
229, 549-558.
(46) Achouba, A.; Dumas, P.; Ouellet, N.; Lemire, M.; Ayotte, P., Plasma levels of selenium-
containing proteins in Inuit adults from Nunavik. Environ. Intern. 2016, 96, 8-15.
(47) Chen, C. Y.; Yu, H. W.; Zhao, J. J.; Li, B.; Qu, L. Y.; Liu, S. P.; Zhang, P. Q.; Chai, Z. F.,
The roles of serum selenium and selenoproteins on mercury toxicity in environmental and
occupational exposure. Environ. Health Perspect. 2006, 114, 297-301.
(48) Palmer, J. H.; Parkin, G., Protolytic cleavage of Hg-C bonds induced by 1-methyl-1,3-
dihydro-2H-benzimidazole-2-selone: Synthesis and structural characterization of mercury
complexes. J. Am. Chem. Soc. 2015, 137, 4503-4516.
(49) Jalilehvand, F.; Leung, B. O.; Izadifard, M.; Damian, E., Mercury(II) cysteine complexes in
alkaline aqueous solution. Inorg. Chem. 2006, 45, 66-73.
(50) Warner, T.; Jalilehvand, F., Formation of Hg(II) tetrathiolate complexes with cysteine at
neutral pH. Can. J. Chem. 2016, 94, 1–7.
(51) Nogara, P. A.; Oliveira, C. S.; Schmitz, G. L.; Piquini, P. C.; Farina, M.; Aschner, M.; Rocha,
J. B. T., Methylmercury's chemistry: From the environment to the mammalian brain. Biochim.
Biophys. Acta 2019, doi.org/10.1016/j.bbagen.2019.01.006.
(52) Manceau, A.; Lemouchi, C.; Enescu, M.; Gaillot, A.-C.; Lanson, M.; Magnin, V.; Glatzel, P.;
Poulin, B. A.; Ryan, J. N.; Aiken, G. R.; Gautier-Luneau, I.; Nagy, K. L., Formation of mercury
sulfide from Hg(II)-thiolate complexes in natural organic matter. Environ. Sci. Technol. 2015, 49,
9787−9796.
(53) Manceau, A.; Lemouchi, C.; Rovezzi, M.; Lanson, M.; Glatzel, P.; Nagy, K. L.; Gautier-
Luneau, I.; Joly, Y.; Enescu, M., Structure, bonding, and stability of mercury complexes with thiolate
and thioether ligands from high-resolution XANES spectroscopy and first-principles calculations.
Inorg. Chem. 2015, 54, 11776−11791.
(54) Manceau, A.; Enescu, M.; Simionovici, A.; Lanson, M.; Gonzalez-Rey, M.; Rovezzi, M.;
Tucoulou, R.; Glatzel, P.; Nagy, K. L.; Bourdineaud, J.-P., Chemical forms of mercury in human hair
reveal sources of exposure. Environ. Sci. Technol. 2016, 50, 10721–10729.
(55) Manceau, A.; Wang, J.; Rovezzi, M.; Glatzel, P.; Feng, X., Biogenesis of mercury-sulfur
nanoparticles in plant leaves from atmospheric gaseous mercury. Environ. Sci. Technol. 2018, 52,
3935–3948.
17
(56) Manceau, A.; Bustamante, P.; Haouz, A.; Bourdineaud, J. P.; Gonzalez-Rey, M.; Geertsen,
V.; Barruet, E.; Rovezzi, M.; Glatzel, P.; Pin, S., Mercury(II) binding to metallothionein in Mytilus
edulis revealed by high energy-resolution XANES spectroscopy. Chem-Eur J. 2019, 25, 997-1009.
(57) Manceau, A.; Nagy, K. L., Thiols in natural organic matter: Molecular forms, acidity, and
reactivity with mercury(II) from First-Principles calculations and high energy-resolution X-ray
absorption near-edge structure spectroscopy. ACS Earth Space Chem. 2019, 3, (12), 2795-2807.
(58) Thomas, S. A.; Catty, P.; Hazemann, J. L.; Michaud-Soret, I.; Gaillard, J. F., The role of
cysteine and sulfide in the interplay between microbial Hg(II) uptake and sulfur metabolism.
Metallomics 2019, 11, (7), 1219-1229.
(59) Bourdineaud, J. P.; Gonzalez-Rey, M.; Rovezzi, M.; Gatzel, P.; Nagy, K. L.; Manceau, A.,
Divalent mercury from dissolved organic matter is bioavailable to fish and accumulates as dithiolate
and tetrathiolate complexes. Environ. Sci. Technol. 2019, 53, 4880−4891.
(60) Bourdineaud, J. P.; Durn, G.; Režun, B.; Manceau, A.; Hrenović, J., The chemical species of
mercury accumulated by Pseudomonas idrijaensis, a bacterium from a rock of the Idrija mercury
mine, Slovenia. Chemosphere 2020, 248, n° 126002.
(61) Thomas, S. A.; Mishra, B.; Myneni, S. C. B., Cellular mercury coordination environment, and
not cell surface ligands, influence bacterial methylmercury production. Environ. Sci. Technol. 2020,
54, 3960-3968.
(62) Manceau, A.; Nagy, K. L.; Glatzel, P.; Bourdineaud, J. P., Molecular and cellular mechanisms
of mercury toxicity to bacteria. Environ. Sci. Technol. 2020, Submitted.
(63) Nagy, K. L.; Manceau, A.; Gasper, J. D.; Ryan, J. N.; Aiken, G. R., Metallothionein-like
multinuclear clusters of mercury(II) and sulfur in peat. Environ. Sci. Technol. 2011, 45, 7298–7306.
(64) George, G. N.; Pickering, I. J.; Pushie, M. J.; Nienaber, K.; Hackett, M. J.; Ascone, I.;
Hedman, B.; Hodgson, K. O.; Aitken, J. B.; Levina, A.; Glover, C.; Lay, P. A., X-ray-induced photo-
chemistry and X-ray absorption spectroscopy of biological samples. J. Synchrotron Radiat. 2012, 19,
875-886.
(65) Rovezzi, M.; Lapras, C.; Manceau, A.; Glatzel, P.; Verbeni, R., High energy-resolution x-ray
spectroscopy at ultra-high dilution with spherically bent crystal analyzers of 0.5 m radius. Rev. Sci.
Instr. 2017, 88, 013108.
(66) Manceau, A.; Marcus, M.; Lenoir, T., Estimating the number of pure chemical components in
a mixture by X-ray absorption spectroscopy. J. Synchrotron Radiat. 2014, 21, 1140-1147.
18
(67) Malinowski, E. R., Theory of error for target factor-analysis with applications to mass-
spectrometry and nuclear magnetic-resonance spectrometry. Anal. Chim. Acta-Comp. 1978, 2, 339-
354.
(68) Manceau, A.; Marcus, M. A.; Tamura, N., Quantitative speciation of heavy metals in soils and
sediments by synchrotron X-ray techniques. In Applications of Synchrotron Radiation in Low-
Temperature Geochemistry and Environmental Science, Fenter, P. A.; Rivers, M. L.; Sturchio, N. C.;
Sutton, S. R., Eds. Mineralogical Society of America: Washington, DC, 2002; Vol. 49, pp 341-428.
(69) Thompson, D. R.; Furness, R. W., The chemical form of mercury stored in South Atlantic
seabirds. Environ. Poll. 1989, 60, (3-4), 305-317.
(70) Cherel, Y.; Barbraud, C.; Lahournat, M.; Jaeger, A.; Jaquemet, S.; Wanless, R. M.; Phillips,
R. A.; Thompson, D. R.; Bustamante, P., Accumulate or eliminate? Seasonal mercury dynamics in
albatrosses, the most contaminated family of birds. Environ. Poll. 2018, 241, 124-135.
(71) McCormack, M. A.; Jackson, B. P.; Dutton, J., Relationship between mercury and selenium
concentrations in tissues from stranded odontocetes in the northern Gulf of Mexico. Sci. Tot. Environ.
2020, 749, n° 141350.
(72) Frodello, J. P.; Romeo, M.; Viale, D., Distribution of mercury in the organs and tissues of five
toothed-whale species of the Mediterranean. Environ. Poll. 2000, 108, 447-452.
(73) Hill, K. E.; Zhou, J. D.; McMahan, W. J.; Motley, A. K.; Atkins, J. F.; Gesteland, R. F.; Burk,
R. F., Deletion of selenoprotein P alters distribution of selenium in the mouse. J. Biol. Chem. 2003,
278, 13640-13646.
(74) Nakayama, A.; Hill, K. E.; Austin, L. M.; Motley, A. K.; Burk, R. F., All regions of mouse
brain are dependent on selenoprotein P for maintenance of selenium. J. Nutr. 2007, 137, (3), 690-
693.
(75) Ralston, N. V. C.; Raymond, L. J., Dietary selenium's protective effects against
methylmercury toxicity. Toxicology 2010, 278, 112-123.
(76) Hill, K. E.; Wu, S.; Motley, A. K.; Stevenson, T. D.; Winfrey, V. P.; Capecchi, M. R.; Atkins,
J. F.; Burk, R. F., Production of selenoprotein P (Sepp1) by hepatocytes is central to selenium
homeostasis. J. Biol. Chem. 2012, 287, 40414-40424.
(77) Quinlivan, P. J. Main Group and Transition Metal Complexes Supported by Carbon, Sulfur,
and Selenium Donor Ligands. Ph.D. Thesis, Columbia University, 2018.
(78) Burk, R. F.; Hill, K. E.; Olson, G. E.; Weeber, E. J.; Motley, A. K.; Winfrey, V. P.; Austin,
L. M., Deletion of apolipoprotein E receptor-2 in mice lowers brain selenium and causes severe
19
neurological dysfunction and death when a low-selenium diet is fed. J. Neurosci. 2007, 27, (23),
6207-6211.
(79) Burk, R. F.; Olson, G. E.; Hill, K. E.; Winfrey, V. P.; Motley, A. K.; Kurokawa, S., Maternal-
fetal transfer of selenium in the mouse. FASEB J. 2013, 27, (8), 3249-3256.
(80) MacDonald, T. C.; Korbas, M.; James, A. K.; Sylvain, N. J.; Hackett, M. J.; Nehzati, S.;
Krone, P. H.; George, G. N.; Pickering, I. J., Interaction of mercury and selenium in the larval stage
zebrafish vertebrate model. Metallomics 2015, 7, (8), 1247-1255.
(81) Earley, J. W., Description and synthesis of the selenide minerals. Am. Miner. 1950, 35, 337-
364.
(82) Pedrosa, L. F. C.; Motley, A. K.; Stevenson, T. D.; Hill, K. E.; Burk, R. F., Fecal selenium
excretion is regulated by dietary selenium intake. Biol. Trace Elem. Res. 2012, 149, 377-381.
(83) Harris, H. H.; Pickering, I. J.; George, G. N., The chemical form of mercury in fish. Science
2003, 301, 1203.
(84) Bustamante, P.; Lahaye, V.; Durnez, C.; Churlaud, C.; Caurant, F., Total and organic Hg
concentrations in cephalopods from the North Eastern Atlantic waters: Influence of geographical
origin and feeding ecology. Sci. Tot. Environ. 2006, 368, (2-3), 585-596.
(85) Kuwabara, J. S.; Arai, Y.; Topping, B. R.; Pickering, I. J.; George, G. N., Mercury speciation
in piscivorous fish from mining-impacted reservoirs. Environ. Sci. Technol. 2007, 41, 2745-2749.
(86) Seco, J.; Xavier, J. C.; Brierley, A. S.; Bustamante, P.; Coelho, J. P.; Gregory, S.; Fielding,
S.; Pardal, M. A.; Pereira, B.; Stowasser, G.; Tarling, G. A.; Pereira, E., Mercury levels in Southern
Ocean squid: Variability over the last decade. Chemosphere 2020, 239, n° 124785.
(87) Torres-Escribano, S.; Velez, D.; Montoro, R., Mercury and methylmercury bioaccessibility in
swordfish. Food. Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 2010, 27, (3), 327-
337.
(88) Clarkson, T. W.; Magos, L., The toxicology of mercury and its chemical compounds. Crit.
Rev. Toxicol. 2006, 36, 609-662.
(89) Bridges, C. C.; Zalups, R. K., Molecular and ionic mimicry and the transport of toxic metals.
Toxicol. Appl. Pharmacol. 2005, 204, 274-308.
(90) Vazquez, M.; Devesa, V.; Velez, D., Characterization of the intestinal absorption of inorganic
mercury in Caco-2 cells. Toxicology in vitro : an international journal published in association with
BIBRA 2015, 29, (1), 93-102.
20
(91) Sherman, L. S.; Blum, J. D.; Franzblau, A.; Basu, N., New insight into biomarkers of human
mercury exposure using naturally occurring mercury stable isotopes. Environ. Sci. Technol. 2013, 47,
3403-3409.
(92) Palmisano, F.; Cardellicchio, N.; Zambonin, P. G., Speciation of mercury in dolphin liver: A
two-stage mechanism for the demethylation accumulation process and role of selenium. Mar.
Environ. Res. 1995, 40, 109-121.
(93) Eagles-Smith, C. A.; Ackerman, J. T.; Yee, J.; Adelsbach, T. L., mercury demethylation in
waterbird livers: dose-response thresholds and differences among species. Environ. Toxicol. Chem.
2009, 28, 568-577.
(94) Wang, Q.; Zhang, C. X.; Ren, Y. S.; Yue, W. B.; Shi, L. G.; Lei, F. L., Molecular structure,
expression analysis and functional characterization of selenoprotein P (SEPP1) in goat (Capra
hircus). J. Anim. Vet. Adv. 2012, 11, 2898-2904.
(95) Korbas, M.; O'Donoghue, J. L.; Watson, G. E.; Pickering, I. J.; Singh, S. P.; Myers, G. J.;
Clarkson, T. W.; George, G. N., The chemical nature of mercury in human brain following poisoning
or environmental exposure. ACS Chem. Neurosci. 2010, 1, 810-818.
(96) Labunskyy, V. M.; Hatfield, D. L.; Gladyshev, V. N., Selenoproteins: Molecular pathways
and physiological roles. Physiol. Rev. 2014, 94, 739-777.
(97) Roy, A.; Kucukural, A.; Zhang, Y., I-TASSER: a unified platform for automated protein
structure and function prediction. Nature Protoc. 2010, 5, 725-738.
(98) Yang, J.; Yan, R.; Roy, A.; Xu, D.; Poisson, J.; Zhang, Y., The I-TASSER Suite: Protein
structure and function prediction. Nat. Methods 2015, 12, 7-8.
(99) Zhang, S. L.; Zang, J. C.; Wang, W. M.; Chen, H.; Zhang, X. R.; Wang, F. D.; Wang, H. F.;
Zhao, G. H., Conversion of the native 24-mer ferritin nanocage into its non-native 16-mer analogue
by insertion of extra amino acid residues. Angew. Chem. Int. Edit. 2016, 55, (52), 16064-16070.
(100) Maity, B.; Abe, S.; Ueno, T., Observation of gold sub-nanocluster nucleation within a
crystalline protein cage. Nat. Commun. 2017, 8, n° 14820.
(101) Pozzi, C.; Ciambellotti, S.; Bernacchioni, C.; Di Pisa, F.; Mangani, S.; Turano, P., Chemistry
at the protein-mineral interface in L-ferritin assists the assembly of a functional (3-oxo)Tris (2-
peroxo) triiron(III) cluster. Proc. Natl. Acad. Sci. U.S.A. 2017, 114, 2580-2585.
(102) Hidalgo, J.; Chung, R. S.; Penkowa, M.; Vasak, M., Structure and function of vertebrate
metallothioneins. Met. Ions Life Sci. 2009, 5, 279–317.
21
(103) Habjanic, J.; Zerbe, O.; Freisinger, E., A histidine-rich Pseudomonas metallothionein with a
disordered tail displays higher binding capacity for cadmium than zinc. Metallomics 2018, 10, 1415-
1429.
(104) Ali, M. A.; Aly, E. M.; Elawady, A. I., Effectiveness of selenium on acrylamide toxicity to
retina. Int. J. Ophthalmol. 2014, 7, (4), 614-620.
(105) DeLano, W. L., The PyMOL Molecular Graphics System. http://www.pymol.org 2002.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI:
Materials, methods, data analysis, supplementary tables and figures (PDF).
HR-XANES spectra (XLS)
AUTHOR INFORMATION
Corresponding Author
E-mail: [email protected]
ORCID
Yves Cherel: 0000-0001-9469-9489
Paco Bustamante : 0000-0003-3877-9390
Pieter Glatzel: 0000-0001-6532-8144
Anne-Claire Gaillot: 0000-0002-2694-6756
Alain Manceau: 0000-0003-0845-611X
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENTS
The authors thank M. Brault-Favrou and C. Churlaud from the Plateforme Analyses Élémentaires of
LIENSs for the trace element analyses. Support was provided to A.M. and P.G. by the ANR under
grant ANR-10-EQPX-27-01 (EcoX Equipex), to P.B. and Y.C. by the Institut Polaire Français Paul
Emile Victor (IPEV programme no. 109, C. Barbraud) and the Terres Australes et Antarctiques
Francaises, and to P.B. by the Institut Universitaire de France (IUF). The AMA and ICP instruments
22
were funded by the Contrat de Plan Etat-Région (CPER) and the European Regional Development
Fund (FEDER). We thank three anonymous reviewers for helpful and constructive comments.
LEGENDS TO FIGURES
Figure 1. Bar charts of the Hg (a) and Se (b) levels in giant petrel tissues, as expressed as a percentage
of the liver levels.
Figure 2. Molar relationship between Hg and Se in the 26 giant petrel tissues analyzed by HR-
XANES and straight line corresponding to the 1:1 ratio. a) Total Se concentration against total Hg
concentration in each tissue. Linear-fit of the liver concentrations (y = -176.1 + 1.83x, R2 = 0.99). b)
Effective Se concentration against total Hg concentration in each tissue. The Y-axis error bars
represent the total propagation of error of the concentrations of Hg and Se and the fractions of Hg
species.
Figure 3. Chemical forms of Hg in tissues of giant petrels derived from Hg L3-edge HR-XANES
spectroscopy. a) Spectra from the brain of individuals P6, P7, and P8 showing the large inter-
individual variability in Hg speciation. b) Spectra from the liver, kidneys, and muscle of the same
individual (P2) with the Hg(Sec)4 spectrum showing the tissular variability in Hg speciation. The
kidneys and muscle spectra are intermediate between the liver (100% HgSe) and the Hg(Sec)4 spectra,
indicating that they contain different amounts of the two species. c-f) Hg species identified by
principal component analysis of 26 tissue spectra and target transformation of HR-XANES spectra
from a library of model compounds and natural species. The MeHgCys spectrum is from Clark’s
grebe feather and the Hg(Sec)4 spectrum is from Clark’s grebe liver.43 The Hg(GSH)2 spectrum is
from Ref.59 The Hg[Me+Cys], Hg[(Sec)4], and Hg[(Cys)2+(N/O)1-2] coordinations of the Hg
complexes are represented in ball-and-stick in the insets and the HgSe structure is represented with
corner-linked Hg(Se)4 tetrahedra. For experimental details and data analysis, see the SI.
Figure 4. Diagrammatic picture of the average concentration and speciation of Hg in giant petrel
tissues. Cys and Sec stand for cysteine and selenocysteine residues within a polymeric chain of
peptide or protein, not for free amino acids. Molar percentages and Hg concentrations are mean values
over three (blood, brain) and five (liver, kidneys, muscle, feathers) individuals. The numbers in
23
parentheses are the standard deviations. HgSe is most abundant in liver and Hg(Sec)4 is most
abundant in muscle.
Figure 5. EDX maps of Hg (green) and Se (red) and transmission electron microscopy images of
HgSe nanoparticles in the liver, kidney and muscle tissues of individual P4. a) EDX map and HRTEM
image of two large HgSe crystals in liver. The upper image on the right is a magnification of the
framed area and the lower image is its fast Fourier transform (FFT) image. The FFT pattern
corresponds to the [112] zone axis of HgSe with a = 6.08 Å. b) EDX map of an aggregate of HgSe
nanoparticles and HRTEM image of three small 4 nm to 10 nm HgSe crystals in kidney. The framed
area is enlarged on the upper right and Fourier transformed on the lower right. c) EDX map of an
aggregate of HgSe nanoparticles in the muscle, and STEM image from 25 drift-corrected individual
frames of two nanocrystals and corresponding FFT images of the upper and lower grains,
respectively. This STEM image was recorded at -170 °C and 60 pA electron current to minimize
radiation damage.
Figure 6. Chemical (a) and effective (b) fractions of Se bound to Hg against total Hg concentration
in each tissue. c) Fraction of bioavailable to total Se against total Hg concentration in each tissue. The
Y-axis error bars represent the total propagation of errors from chemical analysis and HR-XANES
least-squares fits.
Figure 7. Ribbon representation of the predicted tertiary structure from I-Tasser97, 98 of the carboxyl-
terminus region of selenoprotein P for northern fulmar (Fulmarus glacialis), as a proxy for giant
petrel. Purple: Se atoms, blue: N atoms, red: O atoms from the amino acid side chains attached to the
protein backbone in grey. The β-turn forms a molecular cage with four selenolate ligands pointing
out at the surface of the loop. This cage is a potential binding site for Hg. The amino acid sequence
is represented in both single-letter and triple-letter code. U is the single-letter code for selenocysteine.
The structure was visualized with PyMol.105
24
Figure 1
25
Figure 2
26
Figure 3
27
28
Figure 4
29
Figure 5
30
Figure 6
31
Figure 7
S1
Supporting Information for
In vivo formation of HgSe nanoparticles and Hg-tetraselenolate complex from
methylmercury in seabird – Implications for the Hg-Se antagonism
Alain Manceau,*,† Anne-Claire Gaillot,‡ Pieter Glatzel,∥ Yves Cherel,§ Paco Bustamante⊥
†Univ. Grenoble Alpes, CNRS, ISTerre, 38000 Grenoble, France ‡Univ. Nantes, CNRS, Institut des Matériaux Jean Rouxel, IMN, 44000 Nantes, France
∥European Synchrotron Radiation Facility (ESRF), 71 Rue des Martyrs, 38000 Grenoble, France §Centre d’Etudes Biologiques de Chizé (CEBC), CNRS-La Rochelle Université, 79360 Villiers-en-Bois,
France. ⊥La Rochelle Université, CNRS, Littoral Environ. & Soc., LIENSs, 17000 La Rochelle, France
Number of pages: 30
Number of Tables: 3
Number of Figures: 15
Content
S1. Sample preparation
S2. Chemical analysis
S3. X-ray absorption spectroscopy
S3.1. Measurement
S3.2. HR-XANES analysis
S3.3. EXAFS analysis
S4. Electron microscopy
S5. Supplementary tables
S6. Supplementary figures
S7. Supplementary references
S2
S1. Sample preparation
During necropsies, internal tissues (liver, kidneys, pectoral muscle and brain) were sampled, weighed
and wrapped individually in individual plastic bags for trace element analysis. Clotted blood was
collected from heart auricles and stored in microtubes at −20 °C. Body feathers were pulled out from the
lower back and stored dry in individual plastic bags. Birds were sexed during necropsies by visual gonad
examination.
Prior to chemical analyses, internal tissues and blood were freeze-dried, ground to powder and then
stored in plastic vials. Feathers were washed to remove surface dirt and adsorbed contaminants in a
chloroform-methanol solution and then oven dried as described previously.1 For each individual, several
body feathers were pooled to limit potential inter-feather differences in Hg and Se concentrations (n =
20) and were homogenized by cutting them with scissors into small fragments (<1 mm).
S2. Chemical analysis
Total Hg analyses were performed with an AMA-254 mercury analyzer (Altec, Prague, Czech
Republic) on dried tissue aliquots (2–4 mg) by thermal decomposition at 750 °C under oxygen flow with
a detection limit of 0.05 ng Hg. All analyses were repeated 2–3 times until having a standard deviation
<10%. The accuracy was controlled by analysis of a certified reference material (CRM) of lobster
hepatopancreas TORT-2. The TORT-2 recommended value is 0.27 ± 0.06 μg Hg/g dw and the
determined value was 0.264 ± 0.004 μg Hg/g dw (n = 5 measurements).
For Se, 50 to 300 mg tissue aliquots were analyzed using a Thermo Fischer Scientific X series 2 ICP-
MS after digestion with 6 ml 67–70% HNO3 and 2 ml 34–37% HCl (Fisher Scientific, trace element
grade quality). Acid digestion was carried out overnight at room temperature and afterward in a Milestone
microwave digestion system (30 min with a ramping temperature up to 120 °C followed by 15 min
digestion at this temperature). Each sample was then completed to 50 ml with milli-Q water. Three
controls (two CRM and one blank), treated and analyzed in the same way as the samples, were included
in each analytical batch. CRMs were DOLT-5 dogfish liver (NRC, Canada) and TORT-3 (NRC, Canada).
Analyses did not differ from the certified values with a recovery rate of 109 and 104%, respectively. The
detection limit was 0.13 µg/g dry weight (dw). All Hg and Se concentrations are expressed in μg/g dw.
The precision in Hg and Se concentration was assigned to 8% for all measurements.
S3. X-ray absorption spectroscopy
S.3.1. Measurement
All freeze-dried tissues of bird were pressed into 2.5 mm thick pellets, mounted in a polyether ether
ketone (PEEK) sample holder sealed with Kapton tape, and maintained in a desiccator until their transfer
into the liquid helium cryostat of the ID26 beamline. The storage ring was operated in the 7/8 + 1 filling
mode, with 200 mA current. Rejection of higher harmonics and reduction of heat load were achieved
with a white beam Pd-coated, flat mirror working under total reflection at 2.5 mrad deflecting angle. The
energy of the incoming beam was selected by the 111 reflection of a Si double crystal monochromator,
and the beam was focused horizontally by a second Pd-coated mirror and vertically by a third Pd-coated
mirror. The flux on the sample was approximately 1013 photon/s in a beam footprint of ~700 (H) x 80
(V) µm² FWHM. The Hg Lα1 (3d5/2 → 2p3/2) fluorescence line was selected using the 555 reflection of
five spherically bent (radius = 0.5 m)2 Si analyzer crystals (diameter = 100 mm) aligned at 81.8° Bragg
angle in a vertical Rowland geometry. The diffracted intensity was measured with a Si drift diode detector
S3
(SDD) in single photon counting mode. The effective energy resolution, obtained by convoluting the
total instrument energy bandwidth (spreads of the incident and emitted rays) and the 3d5/2 core-hole width
from the Lα1 line was about 3.0 eV, compared to an intrinsic line broadening of about 6.1 eV in
conventional fluorescence yield measurement with a solid-state detector. Spectra were collected at a
temperature of 10-15 K and a scan time of 15 s to reduce exposure, and repeated at different pristine
positions on the sample to increase the signal-to-noise ratio. Scans were monitored carefully for any
evidence of radiation damage. In HR-XANES mode, the incident energy was scanned from 12260 eV to
12360 eV in 0.2 eV steps and the spectra were normalized to unity at E = 12360 eV. In EXAFS mode,
the incident energy was scanned from 12260 eV to 12640 eV in 2.0 eV steps. The stability in energy of
the incident beam was monitored by measuring frequently a fresh MeHgCys reference. The photon
energy is referenced to the maximum of the near-edge peak of MeHgCys at 12279.8 eV. The precision
of the calibrated spectra is 0.1 eV.
S.3.2. HR-XANES analysis
The number of abstract components (i.e., Hg species) in the data set was evaluated in PCA with the
cascade (or variance) of the eigenvalues (EVs),3 the Malinowski’s IND indicator,4, 5 and Malinowski’s
F-test on reduced EVs at 5% significance level.6, 7 The species identities were obtained subsequently by
target transformation8, 9, which makes a linear-squares fit of a reference to a weighted sum of the abstract
components. The degree to which the target transformed spectrum resembles the reference being tested
was evaluated with NSS. The lower the value, the more likely the reference is really represented in the
data set. Lastly, the fractions of the Hg species in each tissue were obtained by fitting linear combinations
of the single-species spectra identified by target transformation to all multicomponent spectra from the
data set. The PCA and linear fit programs from beamline 10.3.2 at the Advanced Light Source were
used.9, 10
Best-fit calculations were conducted initially with one component, then two and three components.
Second and third components were retained only if their addition improved the fit quality, as evaluated
from visual examination of the reconstruction and decrease of NSS by at least 20%. This method has
been used in XANES and EXAFS studies of mercury and other elements such as zinc, iron, and
thallium.11-15 The detection threshold and accuracy of estimation of the fit components derived from the
variation of NSS depends on the quality and complexity of the sample spectrum (i.e., number and nature
of the components), the spectral characteristics of the single-species spectra (i.e., uniqueness of their
features), and how well they represent the unknown sample. A generally accepted rule of thumb is that
the precision on fractional amounts decreases with the number of components. There are exceptions,
however, for example in a ternary system when the two major components have a similar shape and the
third minor component is distinct. A prototypical case is MeHgCys which has an intense near-edge peak
due to the linear coordination of Hg, whereas HgSe and Hg(Sec)4 lack this feature due to the four-
coordination of Hg (Figure 3). Figure S6 shows that it is possible to detect 3% MeHgCys in a mixture
with HgSe and Hg(Sec)4 in the kidneys of petrel P3.
S.3.3. EXAFS analysis
The EXAFS spectra were analyzed with WinXAS16 using amplitude and phase shift functions
generated by FEFF 7.0.217 and β-HgS18 and HgSe19 as structure models. The amplitude reduction factor
S4
S02 was fixed to 0.9. The accuracy of the mean bond distances R and coordination numbers (CN),
including systematic errors, are within 0.02 Å and 20%, respectively (Table S3).
S4. Electron microscopy
Most STEM-HAADF images and elemental maps where acquired at 80 kV and an electron current of
~100 pA to maximize the X-ray fluorescence intensity while preserving at best the organic tissues. In
contrast, the HRTEM images where acquired at 300 kV and ~6 nA to maximize the spatial resolution
while keeping to phase contrast of the strong Hg and Se scatterers.
S5. Supplementary tables
Table S1. Sampling date, location, and biometricsa of northern giant petrels (Macronectes
halli) from Kerguelen Islands.
Individual Sampling date Location Body
mass (kg)
Beak size
(cm)
Tarsus
size (cm)
Wing size
(cm)
P1 18/09/2014 Digby 3.6 9.8 11.3 50.2
P2 12/08/2014 Ratmanov 3.4 11.3 12.2 54
P3 16/09/2014 Ratmanov 4.9 10.2 12.6 53.5
P4 06/07/2014 Pointe Rose 4.9 9.68 12.4 49
P5 29/08/2014 Anse des Papous 5.2 10.2 12 55.5 aThe sampling location and biometrics of individuals 6 to 8 are unknown, but all were southern giant
petrels (M. giganteus).
S5
Table S2. Tissue mercury and selenium concentrations from chemical analysis, Hg speciation from HR-XANES spectroscopy, and bioavailable Se
concentrations and elemental molar ratios used for the bar charts of Figure 1 and the graphs of Figures 2 and 6.
Chemical analysis HR-XANES Hg -bound
Sea
(Hg:Se) effb
(Se:Hg) eff
[Se]bio: [Se]tot
[Se]bio μg/gc
[Se]app μg/gd Petrel Tissue
[Hg] μg/g
[Se] μg/g
(Hg:Se) chem
(Se:Hg) chem
f(MeHgCys) f(Hg(Cys)2) f(HgSe) f(Hg(Sec)4) NSS
P1 Feathers 4.8 5.9 0.32 3.13 100 ± 5 0 0 0 6.2 10-5 0.00 0.00 - 1.00 - -
Liver 226 175 0.51 1.97 < 3 0 100 ± 5 0 ± 4 8.4 10-6 1.00 0.51 1.97 0.49 86.0 86.0
Kidneys 54.5 151 0.14 7.04 < 3 0 62 ± 7 38 ± 7 1.8 10-5 1.00 0.30 3.29 0.70 105.1 129.5
Muscle 26.6 43.3 0.24 4.13 < 3 0 28± 8 72 ± 7 2.9 10-5 0.96 0.77 1.31 0.23 10.2 32.8
Blood 3.2 50.6 0.025 39.80 100 ± 5 0 ± 10 0 0 6.5 10-5 0.00 0.00 - 1.00 50.6 49.3
P2 Feathers 26.2 11.7 0.88 1.14 100 ± 5 0 0 0 4.9 10-5 0.00 0.00 - 1.00 - -
Liver 804 492 0.64 1.55 < 3 0 100 ± 5 0 ± 4 7.8 10-6 1.00 0.64 1.55 0.36 175.5 175.5
Kidneys 37.7 106 0.14 7.14 7 ± 3 0 61 ± 8 32 ± 9 2.2 10-5 0.93 0.26 3.78 0.74 78.0 91.2
Muscle 42.0 50.8 0.33 3.07 < 3 0 33 ± 8 67 ± 8 1.9 10-5 1.00 0.98 1.02 0.02 1.0 34.3
Blood 4.1 97.0 0.017 59.74 n.a. n.a. n.a. n.a. - - - - - - 95.4
P3 Feathers 14.9 6.2 0.95 1.05 100 ± 5 0 0 0 5.4 10-5 0.00 0.00 - 1.00 - -
Liver 293 213 0.54 1.85 < 3 0 91 ± 6 9 ± 6 1.5 10-5 1.00 0.69 1.45 0.31 66.5 97.7
Kidneys 40.9 122 0.13 7.58 12 ± 3 0 62 ± 7 26 ± 9 1.4 10-5 0.88 0.22 4.56 0.78 95.3 105.9
Muscle 21.8 20.9 0.41 2.44 7 ± 3 0 33 ± 9 60 ± 9 2.9 10-5 0.93 1.12 0.89 -0.12 -2.5 12.3
Blood 8.7 102 0.034 29.78 88 ± 8 12 ± 9 0 0 6.4 10-5 0.00 0.00 - 1.00 102.0 98.6
P4 Feathers 12.3 9.9 0.49 2.04 100 ± 5 0 0 0 5.4 10-5 0.00 0.00 - 1.00 - -
Liver 1499 1101 0.54 1.87 < 3 0 90 ± 6 10 ± 6 1.1 10-5 1.00 0.70 1.42 0.30 328.0 510.9
Kidneys 414 363 0.45 2.23 < 3 0 71 ± 7 29 ± 7 1.9 10-5 1.00 0.84 1.19 0.16 58.3 200.0
Muscle 88.7 72.7 0.48 2.08 < 3 0 60 ± 8 40 ± 8 2.2 10-5 1.00 1.06 0.95 -0.06 -4.1 37.8
Blood 4.7 81.9 0.023 44.39 n.a. n.a. n.a. n.a. - - - - - - 80.1
P5 Feathers 10.3 7.8 0.52 1.92 100 ± 5 0 0 0 6.4 10-5 0.00 0.00 - 1.00 - -
Liver 170 109 0.61 1.63 < 3 0 94 ± 6 6 ± 6 1.3 10-5 1.00 0.72 1.38 0.28 30.0 42.1
Kidneys 9.8 55 0.07 14.20 < 3 0 48 ± 9 52 ± 9 3.5 10-5 1.00 0.18 5.55 0.82 45.1 51.1
Muscle 2.9 10.1 0.11 8.96 11 ± 5 0 22 ± 11 67 ± 13 5.0 10-5 0.89 0.32 3.09 0.68 6.8 9.0
Blood 1.3 43.4 0.011 87.50 61 ± 7 39 ± 8 0 0 5.9 10-5 0.00 0.00 - 1.00 43.4 42.9
P6 Feathers 8.5 24.2 0.14 7.23 n.a. n.a. n.a. n.a. - - - - - - -
S6
Liver 214 113 0.74 1.35 n.a. n.a. n.a. n.a. - - - - - - 29.2
Kidneys 27.5 159 0.068 14.70 n.a. n.a. n.a. n.a. - - - - - - 148.5
Muscle 4.2 31.7 0.052 19.36 n.a. n.a. n.a. n.a. - - - - - - 30.1
Brain 2.8 43.1 0.026 38.55 40 ± 6 0 36 ± 13 24 ± 15 4.9 10-5 0.60 0.03 29.21 0.97 41.6 42.0
Blood 3.0 118 0.010 101.02 n.a. n.a. n.a. n.a. - - - - - - 116.9
P7 Feathers 10.3 14.0 0.29 3.45 n.a. n.a. n.a. n.a. - - - - - - -
Liver n.a. n.a. - - n.a. n.a. n.a. n.a. - - - - - - n.a.
Kidneys 39.9 87.1 0.18 5.55 n.a. n.a. n.a. n.a. - - - - - - 71.4
Muscle 1.7 15.1 0.043 23.11 n.a. n.a. n.a. n.a. - - - - - - 14.4
Brain 1.6 26.2 0.024 42.09 83 ± 5 0 8 ± 8 9 ± 9 4.9 10-5 0.17 0.01 95.67 0.99 25.9 25.6
Blood 3.0 68.5 0.017 57.43 n.a. n.a. n.a. n.a. - - - - - - 67.3
P8 Feathers 21.2 8.6 0.97 1.03 n.a. n.a. n.a. n.a. - - - - - - -
Liver 405 191 0.83 1.20 n.a. n.a. n.a. n.a. - - - - - - 32.0
Kidneys 50.8 134 0.15 6.67 n.a. n.a. n.a. n.a. - - - - - - 113.5
Muscle 29.2 35.1 0.33 3.06 n.a. n.a. n.a. n.a. - - - - - - 23.6
Brain 13.2 47.0 0.11 9.05 13 ± 5 0 71 ± 8 16 ± 10 2.0 10-5 0.87 0.15 6.70 0.85 40.0 41.8
Blood 23.9 96.6 0.10 10.26 n.a. n.a. n.a. n.a. - - - - - - 87.2
n.a.: not available. aAtomic fraction. bEffective atomic ratio. cConcentration of bioavailable Se. dApparent concentration of bioavailable Se.
S7
Table S3. Structural parameters derived from the EXAFS analysis of the liver tissue for petrel P2
Hg-Se1 Hg-Hg1 Hg-Se2
CN R (Å) σ2 (Å2) CN R (Å) σ2 (Å2) CN R (Å) σ2 (Å2) ΔE (eV)a
4b 2.64 0.003 12b 4.35 0.016c 12b 5.08 0.016c 5.1 aShift of the energy threshold. bCoordination numbers fixed to crystallographic values.19 cParameters constrained
identical. Fit residual Res = [Σ{|exp – fit|}/Σ{|exp|}] x 100 = 4.7.
S8
S6. Supplementary figures
S9
S10
Figure S1. Plot of the 26 Hg L3-edge HR-XANES spectra measured for this study with their signal-to-
noise ratio (s/n) measured with the two following expressions20:
2
1( / )Var( )
Ss n
S F=
−
2( / )Var(( ) / )
Ss n
S F S=
−
where S and F are the original signal and filtered (i.e., de-noised) data, respectively, is a mean over
the energy range, and Var(x) is the variance of x sampled over the energy range. The de-noising algorithm
is detailed in ref.20.
S11
Figure S2. Mercury L3-edge EXAFS spectra of giant petrel tissues (a) and the tissues from bird P2 only
(b) with the spectrum from Hg(Sec)4.21 The EXAFS spectra of the kidneys and muscles are intermediate
between those of the livers (100% HgSe) and Hg(Sec)4, indicating that they contain variable ratios of
HgSe and Hg(Sec)4.
Figure S3. Estimation by PCA of the number of significant abstract components contained in the 26 HR-
XANES data set from three estimators. a) Scree plot of the components and the associated eigenvalues
(EVs)3. b) Malinowski’s IND indicator4, 5 and Malinowski’s F-test at 5% significance level.6, 7 See ref.20
for details.
S12
Figure S4. Fit of the EXAFS spectrum (a) and Fourier transform (b) of the liver for petrel P2. The
EXAFS parameters are given in Table S3. The Hg-Se1 = 2.64 Å, Hg-Hg1 = 4.35 Å, and Hg-Se2 = 5.08 Å
distances are close to the crystallographic values of 2.63 Å, 4.30 Å, and 5.04 Å for HgSe.19 However,
the Debye-Waller term σ of the Hg-Hg1 and Hg-Se2 pairs, which is the mean-square displacement of the
Hg-Hg1 and Hg-Se2 distances, is extremely large (σ2 = 0.016 Å2). The interatomic distances, although
close on average to those in HgSe, are widely distributed, which means that the nanocrystals are highly
defective, as usually observed in nanoparticles with a high surface/volume ratio.22 The Hg atoms at their
surface have a lower coordination and higher static disorder. Inclusion of sulfur atoms in the first atomic
shell around the Hg atoms (Hg-(S,Se)1 solid-solution) yielded a negative coordination number (CN(S) =
-0.3 and CN(Se) = 4.3), indicating that the HgSe grains do not contain sulfur.
S13
Figure S5. a-b) Low and high magnification STEM-HAADF images of the two large HgSe grains of petrel P4 shown in Figure 5a. The bright
contrast in (b) comes from the large scattering power of Hg and Se. c) EDX maps of Hg (red), Se (green), P (blue), and S (orange). d) Summed
EDX spectrum from the point spectra of the larger grain (framed area #1 in (e)). C, O, P, K, and Fe fluorescence lines are from the organic
tissue and the Cu line from the copper grid. They are not seen in the summed spectrum from the smaller grain sticking out from the tissue (area
#2 – not shown). e) Superimposition of the HAADF image and P, Se and Hg maps. The white arrow across the larger grain indicates the
position and direction of the concentration profiles shown in (f). f) Post-acquisition EDX line profiles (~30 pixels wide = green box in (e))
across the larger grain expressed in atomic fraction (%) of Se (green), Hg (red), P (blue), and S (orange). The black profile is the scattering
contrast of the HAADF image (expressed in kcounts).
S14
Figure S6. a-c) HRTEM images of the same zone shown in Figure S5, demonstrating the crystallinity of the precipitates. Heavy elements (and
thicker areas) appear as dark contrast. c) HRTEM image of the two HgSeNP located in the rectangular region of (a). d,f) Magnification of the
framed area of the smaller (d) and larger (f) HgSe grain in (c). e,g) FFT patterns of the boxed areas in (d) and (f). The patterns correspond to
the [101] (e) and [112] (g) zone axes of HgSe (space group Fd3m) with a = 6.08 Å.
S15
Figure S7. a) STEM-HAADF image of a liver region of petrel P4 rich in HgSeNP aggregates. b) Magnification of the (a) area framed with a
plain line. c) EDX maps in fluorescence counts of Hg (red) and Se (green), and Hg-Se bi-color map of the (b) region. The arrows indicate the
positions and directions of the EDX line profiles 1 and 2 shown in (e). d) Magnification of the (a) area framed with a dashed line (top), and
S16
Hg and Se EDX maps (bottom). The arrow indicates the position and direction of the EDX transect 3, which extends on both sides of the
aggregate. e) Post-acquisition EDX line profiles (~30 pixels wide) of Hg (red) and Se (green) expressed in atomic fraction (%), and scattering
contrast of the HAADF image (black, in kcounts). Transect 3 shows that the area surrounding the HgSeNP aggregate is not enriched in selenium
(e.g., SelP protein). The Se:Hg ratio is statistically equal to 1.
S17
Figure S8. a) TEM image of pieces of dried kidney tissue of petrel P4. b) Zoom of a highly contrasted micron-size aggregate of HgSeNP
corresponding to the boxed area in (a). c) Selected-area electron diffraction (SAED) pattern of the whole aggregate in (b). The first rings match
up with the 111, 220, and 113 diffraction lines of the Fd3m cubic structure of HgSe with a = 6.08 Å. d-f) HRTEM images of isolated HgSeNP
crystals. Star symbols point out less than 3 nm nanoparticules, barely distinguishable on the C-coated membrane (and potentially some organic
matter) of the TEM grid support. e) Zoom of the boxed area in (d) showing the hexagonal phase-contrast pattern compatible with the common
{111} orientation of the cubic structure of NP. The arrow points out a less than 3 nm NP.
S18
Figure S9. a) STEM-HAADF image of several pieces of dried kidney tissue of petrel P4 containing
HgSeNP aggregates of various sizes pointed out with arrows. b) TEM image of the piece framed with a
plain line in (a). The aggregate is >1 µm in length. c) Magnification of the framed area in (b) showing
S19
HgSeNP of different sizes ranging from 5 to 30 nm. d) HAADF image of the large HgSeNP aggregate
framed in the dotted box in (b), and EDX maps of Hg (red), Se (green), and P (bleu) in atomic fraction.
e) Hg-Se-P tri-color map (in atomic fraction) showing the distinct location of P from Hg and Se. The
white arrow indicates the position and direction of the EDX line profile 1 (30 pixels wide) shown in (j).
(f) HAADF image of the HgSeNP aggregate pointed out with the arrow on the upper left of (a). (g)
HAADF image of the area framed with a dotted line on the upper right of (a). h) Magnification of the
framed area in (g) and EDX maps in fluorescence counts of Hg (red) and Se (green). The white arrow
indicates the position and direction of the EDX line profile 2 (30 pixels wide) shown in (j). i) Summed
EDX spectrum from the area framed in yellow in the upper part of image (e). j) Post-acquisition EDX
line profiles 1 and 2 of Hg (red), Se (green), and P (blue) in atomic fraction (%), with the HAADF
contrast (black – in kcounts), showing that the HgSeNP grains are stoichiometric and that P belongs to the
surrounding tissue.
S20
Figure S10. a) STEM-HAADF image of a piece of dried kidney tissue of petrel P4. (b) Magnification of the framed area in (a) showing several
HgSeNP aggregates. The white arrow indicates the position and direction of the EDX line profile 1 (30 pixels wide) displayed in (c). c) Post-
acquisition EDX line profile 1 of the atomic fraction (%) of Hg (red) and Se (green), with HAADF contrast (black, in kcounts), confirming
the HgSe composition of the grains and the absence of Hg and Se in the vicinity of the grains. The spikes are from pixels with low statistic of
S21
X-ray counts in the absence of matter. d) STEM-HAADF image of the two 5 nm aggregates from the framed area in (b) with FFT in (e),
demonstrating their crystalline nature. The white arrow indicates the position and direction of the HAADF intensity profile 2 in (f). (f) Intensity
profile 2 along a 14 nm long line measured below the twinned aggregates showing that the aggregates consist of a mosaic of ~1 nm nanograins.
The nanograins in the twinned aggregates are structurally coherently oriented, as demonstrated by the FFT (e). g) EDX maps of Hg (red) and
Se (green) of the two HgSe twinned aggregates imaged in (b) and (d). The white arrow indicates the position of the EDX line profile 3 displayed
in (h). (h) Post-acquisition EDX line profile 3 of the atomic fraction (%) of Hg (red) and Se (green), with HAADF contrast (black).
S22
Figure S11. a) STEM-HAADF image of dried kidney tissue of petrel P4. b) Zoom on several aggregates of HgSeNP of various sizes ranging
from 5 nm (isolated NP) to 100 nm, and corresponding EDX maps of Hg (red) and Se (green). The arrow indicates the position and direction
of the EDX intensity profile 1 (30 pixels wide) displayed in (c). The areas framed with plain, dashed and dotted lines are zoomed in (d), (e),
and (i), respectively. c)12 pm Post-acquisition EDX line profiles 1 and 2 of Hg (red) and Se (green) in atomic fraction (%), with the HAADF
contrast (black, in kcounts), showing that the HgSeNP grains are stoichiometric throughout the aggregates. d) STEM-HAADF image of the
S23
bottom-left HgSeNP aggregate 2 framed with plain line in (b), and corresponding EDX maps of Hg (red) and Se (green). The arrow indicates
the position and direction of the EDX intensity profile 2 (30 pixels wide) shown in (c). e) HR-STEM-HAADF image of the bottom-right HgSe
aggregate 3 framed with a dashed line in (b). The two plain arrows point out less than 1 or 2 nm size HgSeNP as estimated from the intensity
profile shown in (f). The embedding organic tissue hinders the quality of the image and a better estimate of the NP diameter. f) Intensity profile
3 along the direction shown by the dotted arrow 3 across the NPs in the upper right of (e). g) Zoomed image of the framed area in (e) showing
1-2 nm size NPs pointed out by plain arrows. h) Intensity profile 4 along the direction shown by the dotted arrow 4 across the NPs in (g). i)
HR-STEM-HAADF image of the framed area with dotted line from aggregate 1 in (b), and FFT of the central nanograin. The FFT is consistent
with the Fd3m cubic structure of HgSe with a = 6.08 Å and a [101] zone-axis orientation, as in Fig. S6d,e.
S24
Figure S12. Small aggregates of ~30 nm HgSeNP and isolated 5 to 20 nm HgSeNP were imaged in the
pectoral muscles of petrel P4, owing to the high scattering power of Hg and Se embedded in a thick and
beam-sensitive organic matrix. a) STEM-HAADF image at -170°C of a dried piece of muscle. b)
HAADF image of the region pointed out with an arrow in (a) after degradation of the organic matrix
under the electron beam. The HgSe grain was not altered under the beam, as confirmed by the HR-STEM
contrast in (f). c) Magnification of the right framed area in (a) revealing two HgSeNP (same as those
shown in Fig. 5c). d) EDX maps of Hg (red) and Se (green) from the HgSeNP pointed out in (c). e) Post-
S25
acquisition EDX line profile (position and direction shown in (f)) of the atomic fraction of Hg (red) and
Se (green), and HAADF contrast profile (black, in kcounts), confirming the HgSe composition of the
nanoparticle. f) HR-STEM-HAADF image (5 µs per 17 pm pixel) of the 15 x 20 nm HgSeNP pointed out
in (b). The FFT in inset indicates that the Hg-Se grain has a HgSe cubic structure. g) Intensity periodic
profile along the black arrow (100 pixel wide) of image (f). h) EDX maps of Hg (red) and Se (green) of
the two HgSeNP pointed out in (c). i-j) First (i) and 80th (j) HR-STEM-HAADF images (acquired after
the 7 min long EDX maps), with corresponding FFT patterns, of the two HgSeNP. The twinned
nanoparticles drifted under the 60 pA electron beam, but their crystal structure was preserved. The image
shown in Figure 5c is the sum of the 24 first frames of the time series after correction of the drift artifact.
To prevent damaging the HgSeNP and minimize their displacement during the degradation of the organic
matrix, the total electron dose was drastically reduced by diminishing the counting time per pixel. The
conditions of the acquisitions were 0.2 µs per 12 pm pixel size (instead of 20 µs for other HAADF
images), for a total of 0.1 sec per frame and a cumulated dose of 4733 e-/A2.
S26
Figure S13. a) STEM-HAADF image of a piece of dried pectoral muscle of petrel P4. b) zoom of the framed area in (a). Arrows in (a) and
(b) point out small HgSeNP aggregates. c) STEM-HAADF image of the large HgSe aggregate framed in (b). Arrows 1 and 2 point out two
nanograins. Their size is about 2 nm, as indicated by the HAADF intensity profile shown in inset. d) STEM-HAADF image of isolated HgSeNP
in another region of the sample. e) TEM image of the HgSeNP grains located in the squared area of d). Black arrows point out NP of ~4 nm. f)
EDX maps of Hg (red) and Se (green) of the aggregate in (c). The white arrow indicates the position and direction of the EDX line profile 3
in (f). g) Post-acquisition EDX line profile 3 of Hg (red) and Se (green) in atomic fraction (%), with the HAADF contrast (black – in kcounts),
showing that the HgSeNP are stoichiometric.
S27
Figure S14. a) Total concentration of Se against total concentration of Hg in the liver of giant petrels
P1-P6 and P8 with straight line corresponding to the 1:1 ratio. b) Total and bioavailable concentrations
of Se against total concentration of Hg in the liver of individuals P1-P5. c) Molar proportion of
methylmercury against total Hg concentration in the brain of individuals P6-P8. d) Molar proportions of
HgSe and HgSe+Hg(Sec)4 against total Hg concentration in the brain of individuals P6-P8.
S28
Figure S15. Sensitivity of HR-XANES to MeHgCys mixed with HgSe and Hg(Sec)4 illustrated with the
kidneys spectrum of petrel P3. a) Best-fit calculation with the three Hg species (Table S2). The optimal
molar percentage of MeHgCys is 12%. b) Best-fit calculation without MeHgCys. c) Best-fit calculation
with the percentage of MeHgCys fixed to 9% instead of 12%. The fit residual has increased by 20%
relative to the best-fit value, and the calculated near-edge peak clearly lacks intensity. The spectra on the
right column are an enlargement of the near-edge region.
S29
S7. Supplementary references
(1) Carravieri, A.; Bustamante, P.; Churlaud, C.; Cherel, Y. Penguins as bioindicators of mercury
contamination in the Southern Ocean: Birds from the Kerguelen Islands as a case study. Sci. Tot. Environ.
2013, 454, 141-148.
(2) Rovezzi, M.; Lapras, C.; Manceau, A.; Glatzel, P.; Verbeni, R. High energy-resolution x-ray
spectroscopy at ultra-high dilution with spherically bent crystal analyzers of 0.5 m radius. Rev. Sci. Instr.
2017, 88, 013108.
(3) Cattell, R. B. The scree test for the number of factors. J. Multiv. Behav. Res. 1996, 1, 245–276.
(4) Malinowski, E. R. Determination of the number of factors and the experimental error in a data
matrix. Anal. Chem. 1977, 49, 612-617.
(5) Malinowski, E. R. Factor Analysis in Chemistry. John Wiley and Sons, New York: 2002.
(6) Malinowski, E. R. Statistical F-tests for abstract factor analysis and target testing. J. Chemom. 1988,
3, 49-60.
(7) Malinowski, E. R. Statistical F-tests for abstract factor analysis and target testing: Erratum. J.
Chemom. 1990, 4, 102.
(8) Malinowski, E. R. Theory of error for target factor-analysis with applications to mass-spectrometry
and nuclear magnetic-resonance spectrometry. Anal. Chim. Acta-Comp. 1978, 2, 339-354.
(9) Manceau, A.; Marcus, M. A.; Tamura, N. Quantitative speciation of heavy metals in soils and
sediments by synchrotron X-ray techniques. In Applications of Synchrotron Radiation in Low-
Temperature Geochemistry and Environmental Science, Fenter, P. A.; Rivers, M. L.; Sturchio, N. C.;
Sutton, S. R., EDX. Mineralogical Society of America: Washington, DC, 2002; Vol. 49, pp 341-428.
(10) Marcus, M. A.; MacDowell, A. A.; Celestre, R.; Manceau, A.; Miller, T.; Padmore, H. A.; Sublett,
R. E. Beamline 10.3.2 at ALS: a hard X-ray microprobe for environmental and materials sciences. J.
Synchrotron Radiat. 2004, 11, 239-247.
(11) Isaure, M. P.; Laboudigue, A.; Manceau, A.; Sarret, G.; Tiffreau, C.; Trocellier, P.; Hazemann, J.
L.; Chateigner, D. Quantitative Zn speciation in a contaminated dredged sediment by PIXE, SXRF,
EXAFS spectroscopy and principal component analysis. Geochim. Cosmochim. Acta 2002, 66, 1549-
1567.
(12) Kim, C. S.; Bloom, N. S.; Rytuba, J. J.; Brown, G. E. Mercury speciation by X-ray absorption fine
structure spectroscopy and sequential chemical extractions: A comparison of speciation methods.
Environ. Sci. Technol. 2003, 37, 5102-5108.
(13) Senn, A. C.; Kaegi, R.; Hug, S. J.; Hering, J. G.; Mangold, S.; Voegelin, A. Composition and
structure of Fe(III)-precipitates formed by Fe(II) oxidation in water at near-neutral pH: Interdependent
effects of phosphate, silicate and Ca. Geochim. Cosmochim. Acta 2015, 162, 220-246.
(14) Voegelin, A.; Pfenninger, N.; Petrikis, J.; Majzlan, J.; Plotze, M.; Senn, A. C.; Mangold, S.;
Steininger, R.; Gottlicher, J. Thallium speciation and extractability in a thallium- and arsenic-rich soil
developed from mineralized carbonate rock. Environ. Sci. Technol. 2015, 49, 5390-5398.
(15) Manceau, A.; Wang, J.; Rovezzi, M.; Glatzel, P.; Feng, X. Biogenesis of mercury-sulfur
nanoparticles in plant leaves from atmospheric gaseous mercury. Environ. Sci. Technol. 2018, 52, 3935–
3948.
(16) Ressler, T. WinXAS: a program for X-ray absorption spectroscopy data analysis under MS-
Windows. J. Synchrotron Radiat. 1998, 5, 118-122.
(17) Ankudinov, A. L.; Rehr, J. J. Relativistic calculations of spin-dependent X-ray-absorption spectra.
Phys. Rev. B 1997, 56, 1712-1716.
(18) Rodic, D.; Spasojevic, V.; Bajorek, A.; Onnerud, P. Similarity of structure properties of Hg1-xMnxS
and Cd1-xMnxS (structure properties of HgMnS and CdMnS). J. Mag. Mag. Mater. 1996, 152, 159-164.
(19) Earley, J. W. Description and synthesis of the selenide minerals. Am. Miner. 1950, 35, 337-364.
S30
(20) Manceau, A.; Marcus, M.; Lenoir, T. Estimating the number of pure chemical components in a
mixture by X-ray absorption spectroscopy. J. Synchrotron Radiat. 2014, 21, 1140-1147.
(21) Manceau, A. In preparation. 2020.
(22) Moscoso-Londono, O.; Carriao, M. S.; Cosio-Castaneda, C.; Bilovol, V.; Sanchez, R. M.; Lede, E.
J.; Socolovsky, L. M.; Martinez-Garcia, R. One-step room temperature synthesis of very small -Fe2O3
nanoparticles. Mater. Res. Bull. 2013, 48, 3474-3478.